Department of Hematology/Oncology/Immunology at the Universitätsklinikum of the Philipps-Universität Marburg, Baldingerstrasse, 35043 Marburg, Germany

Correspondence to: A Neubauer, Fax: +49-6421-286-6358

Acute myeloid leukemias (AML) are considered to be clonal disorders involving early hematopoietic progenitor cells. The recent advances in characterization of early stem cells give rise to the question whether it is possible to distinguish healthy progenitors from cells of the leukemic clone in leukemia patients. Differences and similarities in phenotype, genotype and biology are described for leukemic cells and normal hematological progenitors. Recent new insights into human stem cell development offer the perspective that distinction between benign and malignant progenitors might be possible in the future at a very early stage of maturation. Leukemia (2000) 14, 1711-1717.

stem cell; acute myeloid leukemia; CD34; antigen; mutation; cytogenetic aberration

Introduction

For decades, morphology and histochemical staining have been the gold standards for definition of AML according to the FAB (French-American-British Cooperative Group) classification.¹ Immunophenotyping and molecular genetics brought new insights with considerable therapeutic implications into the biology of these diseases. Besides standard chemotherapy hematopoietic stem cell transplantation is an important treatment option for AML. If HLA matched related or unrelated donors are not available, autologous trans- plantation has to be considered as an additional treatment choice. Therefore, distinction between healthy and malignant stem cells has become a major focus of interest and various purging techniques have been proposed in order to improve remission rates and duration.

Current concept of hematopoietic stem cell development

It is a well established model that blood cells derive from pluripotent progenitor cells with the capacity for self-renewal and differentiation. More differentiated cells are generally considered to be more committed to a certain lineage with concomitant loss of self-renewal capacity. Although cell morphology provides good information about the developmental stages of bone marrow cells, immunophenotyping of blood cells is of great help in identifying lineage determination especially in early progenitor cell compartments. Differentiated cells display distinct patterns of antigen expression, like CD11b, CD13, CD33, CD14, CD15 and CD66 antigens on myelomonocytic cells. These lineage markers (lin⁺) are not expressed on early hematopoietic cells. Cells of intermediate differentiation level, which are still capable of giving rise to differentiated progeny, co-express the progenitor cell antigen CD34 and lineage markers. These cells are capable of colony formation in short-term assays termed clonogenic assays such as colony-forming units (CFUs) or burst-forming units (BFUs). In the case of myeloid differentiation they typically carry the CD33 and CD13 antigen.

The leading concept has been that stem cells harbor the CD34 antigen. This led to widespread use of CD34-enriched cell populations in human transplantation. To distinguish the small amount of real stem cells with self-renewal capacity from the majority of CD34-positive cells that have already acquired some degree of differentiation is still the focus of intense interest since the discovery of the CD34 antigen.²

Different model systems are applied to define stem cells with pluripotent hematopoietic potential. Many groups use irradiated non-obese diabetic mice with severe combined immunodeficiency syndrome (NOD/SCID mice) as recipients for human candidate progenitor cells. Others apply in vitro cell culture systems to identify the cells of interest such as long-term culture-initiating cells (LTC-ICs) and limiting dilution assays. The sheep in utero transplantation system seems to be most reliable, because the human progenitor cells are transferred into the sheep fetus in utero before the development of the ovine immune system.^3,4 There is still controversy about suitability and comparability of these model systems for the definition of pluripotent stem cells.⁵

Largely through application of in vitro assays, the surface antigen expression pattern of human hematopoietic stem cells has been defined as CD34⁺ CD38^-. Several other antigens have been described to accompany CD34 expression. Their presence or absence has been reported to define different stem cell populations. The CD45 antigen is expressed on primitive hematopoietic stem cells as well as on all nucleated peripheral blood cells. If the CD34 cells carry the CD90 antigen (Thy-1), they are thought to represent early progenitor cells with superior proliferation and differentiation potential compared to CD90^- cells.^6,7 The same is the case for the AC133 antigen⁸ and the data suggest that the CD82 antigen may also be expressed on stem cells.⁹ Recently, the KDR receptor also known as vascular endothelial growth factor receptor 2 (VEGFR2) has been reported to be a new marker that distinguishes pluripotent hematopoietic stem cells from lineage-committed stem cells.¹⁰ In contrast to this finding, other groups report that VEGFR may define a subtype of hematopoietic progenitor cells with endothelial features.¹¹ The signal-regulatory proteins (SIRPs) have been described very recently as candidate stem cell proteins that are co-expressed on bone marrow CD34⁺, AC133⁺ and CD90⁺ cells.¹² Expression of CD117 antigen (c-kit),¹³ a receptor tyrosine kinase, is low on early human CD34 cells and subsequently becomes more detectable as the maturation process continues.¹⁴ This pattern of antigen expression is also true for HLA-DR¹⁵ and CD38,¹⁶ Flt-3 receptor and interleukin-6 (IL-6) receptor signal transducing element (gp 130),¹⁷ as well as CD71, the transferrin receptor.¹⁸ It varies with the source of progenitor cells.¹⁶

The dogma of the CD34-positive multipotent stem cell has been questioned for about 3 to 4 years. Several groups reported that the earliest hematopoietic progenitor cell has a CD34^-lin^- phenotype^19,20,21 characterized by the CD45 antigen and lack of expression of c-kit, Thy-1 and lineage markers. Expression of CD38 seems to be variable.²² At first sight these findings appear to be in contrast to the established model of CD34⁺ CD38^- phenotype of pluripotent progenitor cells. Other recent publications confirmed the idea of CD34-negative earliest stem cells and present intriguing new data about progenitor cell development, demonstrating the reversibility of CD34 expression on pluripotent hematopoietic cells. This was confirmed in the murine system²³ as well as in humans.²⁴ A new model was consequently proposed: CD34 expression may represent a reversible activation state of hematopoietic stem cells, as revised by Goodell.²⁵

During the past year some groups reported that adult human stem cells have the ability to reconstitute human bone marrow even across the tissue border. They showed that stem cells from the adult brain retained the youthful ability to become several different kinds of tissues. It has been demonstrated that gene-marked adult neural stem cells were able to engraft the bone marrow of sublethally irradiated mice and may thus function as hematopoietic progenitor cells capable of differentiating into myeloid and lymphoid progeny.²⁶ Additionally, non-hematopoietic cells isolated from murine adult skeletal muscle seemed to exhibit even more bone marrow reconstituting hematopoietic activity as compared to whole bone marrow.²⁷ These data evoked controversy of the view that the developmental potential of stem cells is restricted to the differentiated elements of the tissue in which they reside. Figure 1 summarizes these data and hypotheses about normal stem cell phenotype and maturation.

Characterization of the malignant cell clone in acute myeloid leukemia

Acute myeloid leukemia (AML) is thought to be a clonal disorder of poorly differentiated hematopoietic progenitor cells. The current concept of malignancy is based on the idea that altered gene function either confers a growth advantage to a cell or prevents apoptosis. Activation of oncogenes or loss of function of tumor suppressor genes, disturbance of DNA repair genes, alteration of cell cycle genes, etc, results in factor-independent uncontrolled growth and lack of contact inhibition. The idea of a clonal evolution process of malignancy is an important part of that concept, meaning that no single event transforms a cell but additional genetic alterations accumulate and synergize in growth promotion or apotosis inhibition.²⁸ This concept is called 'multistep-carcinogenesis'. Many different kinds of mutations are known to result in altered gene function, eg point mutations, deletions or inversions. These aberrations are often observed in solid tumors. In leukemia and lymphomas, however, gross chromosomal alterations like balanced translocations occur quite frequently, resulting in fusion of unrelated genes and subsequently yielding chimeric proteins with altered functions.

At least 60% of de novo AML harbor cytogenetic abnormalities.²⁹ High-resolution banding techniques suggest that almost all patients have cytogenetic abnormalities.³⁰ Translocations between chromosome 8 and 21, between 15 and 17 and inversions of chromosome 16 have been described most commonly in AML. Quite often AML patients harbor a leukemic clone with multiple cytogenetic aberrations.

During the past years, there has been extensive research on the molecular background of these chromosomal aberrations which elucidated the understanding of leukemogenesis. The t(8;21), which characteristically occurs in FAB M2, results in the AML1-ETO fusion protein. AML1 is the -subunit of core binding factor (CBF) that heterodimerizes with the -subunit and functions as a transcription factor which plays a key role in hematopoiesis. The chimeric protein retains the ability to heterodimerize and bind DNA, but potentially inhibits transcription.^31,32 It also acts as a transactivator for other genes³³ and has been shown to inhibit terminal granulocytic differentiation in cell lines.³⁴

In inversion 16 or less common t(16;16) the -subunit of CBF is altered due to fusion with the tail domain of a smooth muscle myosin heavy chain (SMMYHC), resulting in multimerized complexes of CBF and CBF subunits. These complexes are sequestered preferentially therefore interfering with DNA binding of CBF and its transactivation function.³⁵ Inversion of chromosome 16 is associated with AML FAB subtype M4eo.

The FAB M3 leukemia has a unique translocation pattern, the t(15;17), which results in fusion of the retinoic acid receptor (RAR) to the promyelocytic leukemia gene (PML). DNA binding of this alterated receptor is inappropriate causing a differentiation blockage in myeloid cells.

However, for other less common translocations occurring in AML such as t(3;21), t(12;21), t(9;22), etc, loss or gain of function from genes involved in those lesions have also been described intensively.^{35,36,37,38,39} Excellent reviews on the genes involved in AML with those common translocations providing extensive information about the molecular details and the consecutive biologic effects are presented by Appelbaum et al⁴⁰ and Friedman.³⁵

In addition, there are many reports about gene alterations in AML without any defined gross chromosomal abnormality such as ras-oncogene mutations,^41,42,43 p53 mutations,^44,45 p-glycoprotein expression,⁴⁶ internal tandem dublication of flt3 receptor,^47,48,49 c-myc amplification,^50,51 hox overexpression⁵² or nm23 expression.⁵³ These gene alterations are neither specific for AML FAB subtypes nor do they probably represent causative lesions. But since some typical translocations like t(8;21) and inv 16 may not be sufficient to induce the full malignant potential, these gene mutations may be secondary hits in the process of multistep-carcinogenesis^54,55 (revised in Ref. 35). Some translocations and gene alterations have significant prognostic value indicating a more favorable or fatal outcome.^42,56

Differentiation between healthy and malignant progenitor cells in acute myeloid leukemia

Although extensive research has provided a better understanding of how chromosomal aberrations can disturb gene function resulting in growth advantage or block of differentiation, it is still difficult to distinguish normal from leukemic stem cells. Normal human stem cells and malignant leukemic blast cells exhibit similar morphological, biological and phenotypic features. It is not always possible to distinguish malignant blasts from healthy progenitors solely by morphology. About 20% to 50% of the acute myeloid leukemias show aberrant antigens,⁵⁷ but almost all aberrant antigen combinations can also be detected in a small percentage of healthy individuals. Aberrant antigen expression may also vary during the course of AML treatment and is therefore not suitable as a single reliable marker of the malignant cell clone.

In the absence of aberrant antigen combinations, malignant myeloid cells in AML often display a very similar surface protein expression pattern as compared with their healthy CD34 counterpart. Until now there has been almost no antigen described that has proved to be expressed on malignant blast cells exclusively. The AC133 antigen is neither specific for non-AML blast cells nor for myeloid blast cells.^58,59 The CD82 antigen is also expressed on hematopoietic progenitor cells and on AML cells.⁹ The Flt3-receptor tyrosine kinase, which is up-regulated on normal activated stem cells, has high m-RNA and protein expression levels in AML cells.⁶⁰ However, the internal tandem duplication within the juxtamembrane region of that receptor occurs exclusively in AML cells. It has been reported that the expression pattern of adhesion molecules is different in cells from AML patients and normal stem cells, but the differences are quantitative and not absolutely specific.⁶¹ Whether the KDR receptor also occurs on myeloid leukemia cells has not been investigated until now. The human signal-regulatory protein (SIRP), described as a stem cell marker very recently, is a serious candidate protein for distinction between hematopoietic progenitor cells and myeloid blasts because the latter cells do not express that protein or express it in a significantly reduced manner.¹² Although the malignant leukemic blasts in AML generally arise from very early CD34⁺ progenitor cells, the FAB subtypes of acute myeloid leukemia seem to have different differentiation levels. The CD13 and CD33 antigens and myeloperoxidase that characterize normal myeloid progenitor cells are abundantly expressed on almost all AML cells, but are variable in FAB M0, which is considered to be a rather undifferentiated leukemia. Most of the leukemic cells of this FAB subtype exhibit a CD34⁺ CD38^- antigen expression and lack other progenitor cell antigens.⁶² The c-kit antigen expression is present in about 60% of AML in general but differs between FAB subtypes and karyotype, AML M0 and M1 leukemias show more pronounced c-kit-antigen expression as compared with M5 leukemias.⁶³ The M2 leukemia which frequently harbors the t(8;21) and even the promyeolocytic FAB M3 leukemia with t(15;17) have recently been characterized as diseases of multipotent progenitor cells with lineage-negative phenotype (CD34⁺ CD38^-).^64,65,66 AML M4 and M5 leukemias are thought to display a more mature phenotype often carrying a lineage-positive monocytic antigen pattern. Lineage restriction may also occur in AML M6 and M7, which is underscored by morphology and flow cytometry (eg glycophorin and glycoprotein IIb/IIIa expression, respectively). Nevertheless, Dicks group demonstrated the immature nature of acute myeloid leukemia cells,⁶⁷ and Blair et al⁶⁸ characterized AML cells as CD34⁺ HLA-DR^- CD71^- CD90^-15 which represents an early progenitor cell phenotype, however, most of their examined AML cases were M4 and M5 leukemias.

There is one report on differentiative and proliferative capacities of leukemic blast, that demonstrates the potential for self-renewal of leukemic stem cells in different AML subtypes.⁶⁹ This study strongly supports the hypothesis of a primitive stem cell being involved in acute leukemic transformation, although heterogeneous maturation characteristics were observed in different AML FAB subtypes. Bonnet and Dick⁶⁹ proved that all cells capable of initiating human AML in NOD-SCID mice were exclusively CD34⁺CD38^-. They also demonstrated that some features of cell growth remain similar in normal and malignant blast cells.

According to the data of Blair's group CD34⁺ CD90⁺ progenitor cells were thought to be benign cells in AML. In order to analyze this particular antigen pattern in AML in detail, we used multiparameter flow cytometry sorting, cytogenetics and FISH techniques. To this end, we have shown that in secondary AML the early CD34⁺ CD90⁺ cells also carry the malignant clone,⁷⁰ meaning that in this disease leukemic cells resemble the earliest CD34 cell progenitor phenotype. The patients examined in this study had progressive disease with poor response to chemotherapy treatment. The discordant findings could be explained with the different study populations (primary vs secondary AML, different FAB subtypes). It is also very likely that the malignant cell clone predominates the earliest stem cell compartments only in later or progressive stages as is the case in chronic myeloid leukemia.⁷¹ However, this controversy reflects the difficulty of defining early and more mature acute myeloid leukemia cases solely by immunophenotype. Acute myeloid leukemias that predominantly display a CD34^- CD33⁺ phenotype might also harbor early malignant clones within the CD34⁺ CD90⁺ population. This clone will probably expand during disease progression due to loss of the ability to differentiate further or some growth advantage after chemotherapy treatment. In CML it has been demonstrated that the early malignant bcr-abl-positive CD34⁺ cells stimulate their growth through autocrine IL-3 and G-CSF release. These cells lose this capacity during maturation and differentiation.⁷² A similar mechanism could be assumed for AML cells.

The discordance of our findings as compared with the literature can also be explained in part by different methods applied while addressing the question of developmental stages of stem cells. Most investigators make use of karyotypic abnormalities in AML cells in order to identify the malignant clone applying either conventional G-banding cytogenetic analysis, interphase fluorescence in situ hybridization (FISH) analysis or PCR-based molecular genetic techniques. Several pitfalls have been demonstrated for these techniques. In certain cytogenetic lesions such as monosomy 7 or t(15;17) standard G-banding analysis obscures the high number of aberrant karyotypes within blood cell fractions, probably due to different proliferative abilities of the malignant clone in vitro.^73,74 One could assume that with other leukemic karyotypic lesions false positive results could be obtained with the standard cytogenetic analysis because of higher proliferative potential of the malignant cell type in cell culture compared to healthy blood cells. In some specific lesions, interphase FISH seems to be superior in detecting the cytogenetic aberration. Although FISH is a widely accepted sensitive standard technique for the detection of specific cytogenetic aberrations in semi-quantitative manner, false positive signals can appear due to artifacts when analyzing three-dimensional cells from a two-dimensional perspective. Therefore, cut-off levels have to be defined for each probe on healthy individuals. If the cell of interest has been exposed to stressful events such as sorting, culturing etc, cut-off levels may be significantly higher and may vary considerably due to a higher amount of dead cells within a fraction or squeezed nuclei resulting in interpretation failure.⁷⁵ To avoid these difficulties, polymerase chain reaction (PCR) analysis can be applied as an equal or even superior technique. It is a well established method for sensitive detection of specific gene rearrangements such as bcr-abl, t(15;17) or t(8;21). For detection of inversion 16 it seems to be more reliable than standard cytogenetic analysis.⁷⁶ Despite the high sensitivity, standard PCR analysis is not suitable for exact quantification. New technologies for REAL-time PCR analysis may partially overcome that problem. However, only defined genetic alterations can be sensitively detected by PCR and FISH technique and require well defined probes that are available only for the most common genetic lesions. Since AML patients display a variety of chromosomal defects, which are often not detectable either with FISH or PCR analysis, combination of both techniques with standard cytogenetic analysis is recommended.

Different genetic alterations within leukemic cells seem to have very unique biological features reflecting the heterogeneity of AML cases. All patients whose blasts harbor t(8;21) and inv16 (CBF⁺ leukemias) have a favorable outcome.^77,78 One could speculate that in those leukemias the cells are more susceptible to chemotherapy but there is no proven explanation for the benefit of those rearrangements. Promyelocytes in t(15;17) M3 leukemia respond to retinoic acid treatment with differentiation. This treatment improves survival in AML patients with this particular lesion. Another type of acute myeloid leukemia, familial AML with monosomy 7, has been shown to affect very early multipotent progenitor cells and to have extraordinarily poor patient survival.⁷⁹ In addition, AML patients with a history of myelodysplastic syndrome and monosomy 7 often have a fatal outcome. One could assume that chromosome 7 harbors a gene important in apoptotic pathways but until now such a target gene has not been identified. Another example of biological significance of certain genetic lesions is the Flt3 gene. Its alteration can modify the proliferative ability of leukemic cells.⁴⁷

As mentioned above, biological features of leukemic and normal cells in AML are often similar.⁶⁹ Leukemic cells can form short-term colonies and have long-term proliferative potential in vitro.¹⁵ Similar to normal CD34⁺ progenitors some AML cells can be forced to differentiate into dendritic cells⁸⁰ and can also be stimulated to differentiate according to their lineage, for example with CD44 antibody in M1/2 to M5 leukemia.⁸¹

To date, it can be stated that combined analyses of stem cell phenotypic antigen expression, genetic lesions and biological features in AML provide useful tools to detect differences between normal and malignant progenitor cells that are of clinical benefit possibly providing data for new sorting strategies in this disease.

Perspectives

It has been reported that in chronic myeloid leukemias (CML) or myelodysplastic syndrome (MDS) sorting of benign stem cells with a certain phenotype could be accomplished.^71,82 These diseases are also considered to be clonal disorders of multipotent progenitor cells.⁸³ But to date for autologous transplantation, no reliable purging method exists for obtaining leukemia free grafts.

The current model system of CD34^- phenotype of the earliest progenitor cell and switch of the resting stem cell from CD34^- to CD34⁺ phenotype due to external activation stimuli gives rise to the question whether this cell might be free from leukemic genetic lesions providing a source for healthy stem cells in autologous transplantation. This hypothesis seems likely, because the very early progenitor cell is probably a dormant cell, less prone to gene altering events that can occur in proliferating cells. If chemotherapy agents or radiation alter the recovery state of the bone marrow, the stem cells attain an activation state upon external proliferation stimuli. One could imagine that through activation the prior dormant stem cells respond with multiple surface receptor up-regulation and therefore become CD34⁺ cells. If some of these activated CD34⁺ cells enter the cell cycle in order to reconstitute the bone marrow cell pool they become susceptible targets for DNA damaging agents such as benzene or alkylating substances. These cells are more prone to becoming leukemic than dormant cells. The effort of detecting a leukemic-free cell compartment within the CD34⁺ population might therefore be frustrating.

In addition to their dormant state, CD34^- hematopoietic stem cells possibly have potent mechanisms protecting them from genetic damage like efflux pumps for toxic agents. Therefore, the dye efflux definition of CD34^- stem cells seems likely.²² The expectation that these cells might display a compartment free of leukemic cell clones is very promising but needs to be proven in future studies. If these early progenitors could be identified and mobilized in vivo for collection, only very few cells might be sufficient for engraftment in stem cell transplantation settings and might provide a new source of leukemic-free graft with reduced risk of leukemic relapse. According to that model, CD34^- cells will turn into CD34⁺ cells that up-regulate growth factor receptors, kit-ligand and other surface antigens in response to external stimuli, resulting in augmented self-renewal and differentiation potential.

This new perspective for autologous bone marrow transplantation without risk of malignant cell contamination becomes even more likely if one considers the data about bone marrow repopulating potential of non-hematologic stem cells from brain or muscle tissue. It is not expected that these tissues will be affected by the malignant leukemic cell clone thus possibly providing a safe source for leukemic-free bone marrow reconstituting cells useful for stem cell trans- plantation. We assume that these exciting new data will challenge scientists and physicians to co-operate in further investigations in the area of stem cell characterization.

Acknowledgements

This work was supported in part by grants of the Deutsche Forschungsgemeinschaft (Ne 310/6-3), and the José Carreras Stiftung.

1 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gralnick HR, Sultan C. Proposals for the classification of the acute leukaemias. French-American-British (FAB) co-operative group. Br J Haematol 1976; 33: 451-458, MEDLINE

2 Civin CI, Strauss LC, Brovall C, Fackler MJ, Schwartz JF, Shaper JH. Antigenic analysis of hematopoiesis. III. A hematopoietic progenitor cell surface antigen defined by a monoclonal antibody raised against KG- 1a cells. J Immunol 1984; 133: 157-165, MEDLINE

3 Zanjani ED, Almeida-Porada G, Livingston AG, Flake AW, Ogawa M. Human bone marrow CD34- cells engraft in vivo and undergo multilineage expression that includes giving rise to CD34+ cells (see comments). Exp Hematol 1998; 26: 353-360, MEDLINE

4 Zanjani ED, Almeida-Porada G, Livingston AG, Porada CD, Ogawa M. Engraftment and multilineage expression of human bone marrow CD34- cells in vivo. Ann NY Acad Sci 1999; 872: 220-231; discussion 231-232, MEDLINE

5 Larochelle A, Vormoor J, Hanenberg H, Wang JC, Bhatia M, Lapidot T, Moritz T, Murdoch B, Xiao XL, Kato I, Williams DA, Dick JE. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: implications for gene therapy. Nature Med 1996; 2: 1329-1337, MEDLINE

6 Murray L, Chen B, Galy A, Chen S, Tushinski R, Uchida N, Negrin R, Tricot G, Jagannath S, Vesole D, Barlogie B, Hoffman R, Tsukamoto A. Enrichment of human hematopoietic stem cell activity in the CD34+Thy- 1+Lin- subpopulation from mobilized peripheral blood. Blood 1995; 85: 368-378, MEDLINE

7 Humeau L, Bardin F, Maroc C, Alario T, Galindo R, Mannoni P, Chabannon C. Phenotypic, molecular, and functional characterization of human peripheral blood CD34+/THY1+ cells. Blood 1996; 87: 949-955, MEDLINE

8 de Wynter EA, Buck D, Hart C, Heywood R, Coutinho LH, Clayton A, Rafferty JA, Burt D, Guenechea G, Bueren JA, Gagen D, Fairbairn LJ, Lord BI, Testa NG. CD34+AC133+ cells isolated from cord blood are highly enriched in long-term culture-initiating cells, NOD/SCID-repopulating cells and dendritic cell progenitors. Stem Cells 1998; 16: 387-396, MEDLINE

9 Burchert A, Notter M, Menssen HD, Schwartz S, Knauf W, Neubauer A, Thiel E. CD82 (KAI1), a member of the tetraspan family, is expressed on early haemopoietic progenitor cells and up-regulated in distinct human leukaemias. Br J Haematol 1999; 107: 494-504, MEDLINE

10 Ziegler BL, Valtieri M, Porada GA, De Maria R, Muller R, Masella B, Gabbianelli M, Casella I, Pelosi E, Bock T, Zanjani ED, Peschle C. KDR receptor: a key marker defining hematopoietic stem cells. Science 1999; 285: 1553-1558, Article MEDLINE

11 Gehling UM, Ergun S, Schumacher U, Wagener C, Pantel K, Otte M, Schuch G, Schafhausen P, Mende T, Kilic N, Kluge K, Schafer B, Hossfeld DK, Fiedler W. In vitro differentiation of endothelial cells from AC133-positive progenitor cells (in process citation). Blood 2000; 95: 3106-3112, MEDLINE

12 Seiffert M, Cant C, Chen Z, Rappold I, Brugger W, Kanz L, Brown EJ, Ullrich A, HJ Bh. Human signal-regulatory protein is expressed on normal, but not on subsets of leukemic myeloid cells and mediates cellular adhesion involving Its counterreceptor CD47. Blood 1999; 94: 3633-3643, MEDLINE

13 Ashman LK, Cambareri AC, Levinsky RJ, Juttner CA. Expression of the YB5.B8 antigen (c-kit protooncogene product) in normal human bone marrow. Blood 1991; 78: 30-37, MEDLINE

14 Sogo S, Inaba M, Ogata H, Hisha H, Adachi Y, Mori S, Toki J, Yamanishi K, Kanzaki H, Adachi M, Ikehara S. Induction of c-kit molecules on human CD34+/c-kit < low cells: evidence for CD34+/c-kit < low cells as primitive hematopoietic stem cells. Stem Cells 1997; 15: 420-429, MEDLINE

15 Blair A, Hogge DE, Sutherland HJ. Most acute myeloid leukemia progenitor cells with long-term proliferative ability in vitro and in vivo have the phenotype CD34(+)/CD71(-)/HLA-DR. Blood 1998; 92: 4325-4335, MEDLINE

16 Roy V, Miller JS, Verfaillie CM. Phenotypic and functional characterization of committed and primitive myeloid and lymphoid hematopoietic precursors in human fetal liver. Exp Hematol 1997; 25: 387-394, MEDLINE

17 Kollet O, Aviram R, Chebath J, Ben-Hur H, Nagler A, Shultz L, Revel M, Lapidot T. The soluble interleukin-6 (IL-6) receptor/IL-6 fusion protein enhances in vitro maintenance and proliferation of human CD34(+)CD38(-/low) cells capable of repopulating severe combined immunodeficiency mice. Blood 1999; 94: 923-931, MEDLINE

18 Gross S, Helm K, Gruntmeir JJ, Stillman WS, Pyatt DW, Irons RD. Characterization and phenotypic analysis of differentiating CD34+ human bone marrow cells in liquid culture. Eur J Haematol 1997; 59: 318-326, MEDLINE

19 Bhatia M, Bonnet D, Murdoch B, Gan OI, Dick JE. A newly discovered class of human hematopoietic cells with SCID-repopulating activity (see comments). Nature Med 1998; 4: 1038-1045, Article MEDLINE

20 Osawa M, Hanada K, Hamada H, Nakauchi H. Long-term lymphohematopoietic reconstitution by a single CD34- low/negative hematopoietic stem cell. Science 1996; 273: 242-245, MEDLINE

21 Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med 1996; 183: 1797-1806, MEDLINE

22 Goodell MA, Rosenzweig M, Kim H, Marks DF, DeMaria M, Paradis G, Grupp SA, Sieff CA, Mulligan RC, Johnson RP. Dye efflux studies suggest that hematopoietic stem cells expressing low or undetectable levels of CD34 antigen exist in multiple species. Nature Med 1997; 3: 1337-1345, MEDLINE

23 Sato T, Laver JH, Ogawa M. Reversible expression of CD34 by murine hematopoietic stem cells (see comments). Blood 1999; 94: 2548-2554, MEDLINE

24 Nakamura Y, Ando K, Chargui J, Kawada H, Sato T, Tsuji T, Hotta T, Kato S. Ex vivo generation of CD34(+) cells from CD34(-) hematopoietic cells. Blood 1999; 94: 4053-4059, MEDLINE

25 Goodell MA. Introduction: focus on hematology. CD34(+) or CD34(-): does it really matter? (comment). Blood 1999; 94: 2545-2547, MEDLINE

26 Bjornson CR, Rietze RL, Reynolds BA, Magli MC, Vescovi AL. Turning brain into blood: a hematopoietic fate adopted by adult neural stem cells in vivo (see comments). Science 1999; 283: 534-537, Article MEDLINE

27 Jackson KA, Mi T, Goodell MA. Hematopoietic potential of stem cells isolated from murine skeletal muscle (in process citation). Proc Natl Acad Sci USA 1999; 96: 14482-14486, Article MEDLINE

28 Loeb LA. Cancer cells exhibit a mutator phenotype. Adv Cancer Res 1998; 72: 25-56, MEDLINE

29 Grimwade D, Walker H, Oliver F, Wheatley K, Harrison C, Harrison G, Rees J, Hann I, Stevens R, Burnett A, Goldstone A. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. Blood 1998; 92: 2322-2333, MEDLINE

30 Yunis JJ, Brunning RD, Howe RB, Lobell M. High-resolution chromosomes as an independent prognostic indicator in adult acute nonlymphocytic leukemia. New Engl J Med 1984; 311: 812-818, MEDLINE

31 Frank R, Zhang J, Uchida H, Meyers S, Hiebert SW, Nimer SD. The AML1/ETO fusion protein blocks transactivation of the GM-CSF promoter by AML 1B. Oncogene 1995; 11: 2667-2674, MEDLINE

32 Meyers S, Lenny N, Hiebert SW. The t(8;21) fusion protein interferes with AML-1B-dependent transcriptional activation. Mol Cell Biol 1995; 15: 1974-1984, MEDLINE

33 Rhoades KL, Hetherington CJ, Rowley JD, Hiebert SW, Nucifora G, Tenen DG, Zhang DE. Synergistic up-regulation of the myeloid-specific promoter for the macrophage colony-stimulating factor receptor by AML1 and the t(8;21) fusion protein may contribute to leukemogenesis. Proc Natl Acad Sci USA 1996; 93: 11895-11900, Article MEDLINE

34 Ahn MY, Huang G, Bae SC, Wee HJ, Kim WY, Ito Y. Negative regulation of granulocytic differentiation in the myeloid precursor cell line 32Dc13 by ear-2, a mammalian homolog of Drosophila seven-up, and a chimeric leukemogenic gene, AML1/ETO. Proc Natl Acad Sci USA 1998; 95: 1812-1817, MEDLINE

35 Friedman AD. Leukemogenesis by CBF oncoproteins. Leukemia 1999; 13: 1932-1942, MEDLINE

36 Zent C, Rowley JD, Nucifora G. Rearrangements of the AML1/CBFA2 gene in myeloid leukemia with the 3;21 translocation: in vitro and in vivo studies. Leukemia 1997; 11: (Suppl. 3) 273-278, MEDLINE

37 Rubnitz JE, Pui CH, Downing JR. The role of TEL fusion genes in pediatric leukemias. Leukemia 1999; 13: 6-13, MEDLINE

38 Look AT. Oncogenic transcription factors in the human acute leukemias. Science 1997; 278: 1059-1064, MEDLINE

39 Lim LC, Heng KK, Vellupillai M, Tan LT, Boey BC, Lau LC, How GF. Molecular and phenotypic spectrum of de novo Philadelphia positive acute leukemia. Int J Mol Med 1999; 4: 665-667, MEDLINE

40 Appelbaum FR, Gilliland DG, Tallmann MS. The biology and treatment of acute myeloid leukemia. American Society of Hematology Education Program Book 1998; pp 15-43,

41 Pedersen-Bjergaard J, Janssen JWG, Lyons J, Philip P, Bartram CR. Point mutations of the ras protooncogenes and chromosome aberrations on acute nonlymphoblastic leukemia and preleukemia related to therapy with alkylating agents. Cancer Res 1988; 48: 1812-1817, MEDLINE

42 Neubauer A, Dodge R, George SL, Davey FR, Silver RT, Schiffer CA, Mayer R, Ball ED, Wurster-Hill D, Bloomfield CD, Liu ET. Prognostic importance of mutations in the ras protooncogenes in de novo acute myeloid leukemia. Blood 1994; 83: 1603-1611, MEDLINE

43 Schmidt CA, Przybylski G, Vogel D, Ludwig WD, Oettle H, Neubauer A, Siegert W. ras point mutations occur in acute myeloid leukemia with illegitimate T-cell receptor delta gene rearrangement. Leukemia 1994; 8: 102-105, MEDLINE

44 Seliger B, Papadileris S, Vogel D, Hess G, Brendel C, Storkel S, Oertel J, Kolbe K, Huber C, Huhn D, Neubauer A. Analysis of p53 and mdm-2 gene in acute myeloid leukemia. Eur J Haematol 1996; 57: 230-240, MEDLINE

45 Slingerland JM, Minden MD, Benchimol S. Mutation of the p53 gene in human acute myelogeneous leukemia. Blood 1991; 77: 1500-1507, MEDLINE

46 Marie JP, Legrand O. MDR1/P-GP expression as a prognostic factor in acute leukemias. Adv Exp Med Biol 1999; 457: 1-9, MEDLINE

47 Rombouts WJ, Broyl A, Martens AC, Slater R, Ploemacher RE. Human acute myeloid leukemia cells with internal tandem duplications in the Flt3 gene show reduced proliferative ability in stroma supported long-term cultures. Leukemia 1999; 13: 1071-1078, MEDLINE

48 Kiyoi H, Naoe T, Nakano Y, Yokota S, Minami S, Miyawaki S, Asou N, Kuriyama K, Jinnai I, Shimazaki C, Akiyama H, Saito K, Oh H, Motoji T, Omoto E, Saito H, Ohno R, Ueda R. Prognostic implication of FLT3 and N-RAS gene mutations in acute myeloid leukemia. Blood 1999; 93: 3074-3080, MEDLINE

49 Xu F, Taki T, Yang HW, Hanada R, Hongo T, Ohnishi H, Kobayashi M, Bessho F, Yanagisawa M, Hayashi Y. Tandem duplication of the FLT3 gene is found in acute lymphoblastic leukaemia as well as acute myeloid leukaemia but not in myelodysplastic syndrome or juvenile chronic myelogenous leukaemia in children. Br J Haematol 1999; 105: 155-162, MEDLINE

50 Alitalo K, Saksela K, Winqvist R, Alitalo R, Keski-Oja J, Laiho M, Ilvonen M, Knuutila S, Delachapelle A. Acute myelogenous leukemia with c-myc amplification and double minute chromosome. Lancet 1985; 11: 1035-1038,

51 Baer MR, Augustinos P, Kinniburgh AJ. Defective c-myc and c-myb RNA turnover in acute myeloid leukemia cells. Blood 1992; 79: 1319-1326, MEDLINE

52 Thorsteinsdottir U, Sauvageau G, Hough MR, Dragowska W, Lansdorp PM, Lawrence HJ, Largman C, Humphries RK. Overexpression of HOXA10 in murine hematopoietic cells perturbs both myeloid and lymphoid differentiation and leads to acute myeloid leukemia. Mol Cell Biol 1997; 17: 495-505, MEDLINE

53 Okabe-Kado J, Kasukabe T, Honma Y. Differentiation inhibitory factor Nm23 as a prognostic factor for acute myeloid leukemia. Leuk Lymphoma 1998; 32: 19-28, MEDLINE

54 Okuda T, Cai Z, Yang S, Lenny N, Lyu CJ, van Deursen JM, Harada H, Downing JR. Expression of a knocked-in AML1-ETO leukemia gene inhibits the establishment of normal definitive hematopoiesis and directly generates dysplastic hematopoietic progenitors. Blood 1998; 91: 3134-3143, MEDLINE

55 Castilla L, Bodine D, Garrett L, Liu P. Leukemia development by the induction of a second 'hit' in chimeric mice containing CBF-beta-MYH11 ES cells. Blood 1998; 92: 214a,

56 Bloomfield CD, Lawrence D, Byrd JC, Carroll A, Pettenati MJ, Tantravahi R, Patil SR, Davey FR, Berg DT, Schiffer CA, Arthur DC, Mayer RJ. Frequency of prolonged remission duration after high-dose cytarabine intensification in acute myeloid leukemia varies by cytogenetic subtype. Cancer Res 1998; 58: 4173-4179, MEDLINE

57 DeVita VTJ, Hellman S, Rosenberg SA. Cancer - Principles and Practice of Oncology, 4th edn. JB Lippincott Co: Philidelphia, 1993,

58 Buhring HJ, Seiffert M, Marxer A, Weiss B, Faul C, Kanz L, Brugger W. AC133 antigen expression is not restricted to acute myeloid leukemia blasts but is also found on acute lymphoid leukemia blasts and on a subset of CD34+ B-cell precursors (letter). Blood 1999; 94: 832-833, MEDLINE

59 Horn PA, Tesch H, Staib P, Kube D, Diehl V, Voliotis D. Expression of AC133, a novel hematopoietic precursor antigen, on acute myeloid leukemia cells (letter). Blood 1999; 93: 1435-1437, MEDLINE

60 Drexler HG, Meyer C, Quentmeier H. Effects of FLT3 ligand on proliferation and survival of myeloid leukemia cells. Leuk Lymphoma 1999; 33: 83-91, MEDLINE

61 De Waele M, Renmans W, Jochmans K, Schots R, Lacor P, Trullemans F, Otten J, Balduck N, Vander Gucht K, Van Camp B, Van Riet I. Different expression of adhesion molecules on CD34+ cells in AML and B- lineage ALL and their normal bone marrow counterparts. Eur J Haematol 1999; 63: 192-201, MEDLINE

62 Costello R, Mallet F, Chambost H, Sainty D, Arnoulet C, Gastaut JA, Olive D. The immunophenotype of minimally differentiated acute myeloid leukemia (AML-MO): reduced immunogenicity and high frequency of CD34+/CD38- leukemic progenitors. Leukemia 1999; 13: 1513-1518, MEDLINE

63 Schwartz S, Heinecke A, Zimmermann M, Creutzig U, Schoch C, Harbott J, Fonatsch C, Loffler H, Buchner T, Ludwig WD, Thiel E. Expression of the C-kit receptor (CD117) is a feature of almost all subtypes of de novo acute myeloblastic leukemia (AML), including cytogenetically good-risk AML, and lacks prognostic significance. Leuk Lymphoma 1999; 34: 85-94, MEDLINE

64 Wittebol S, Raymakers R, van de Locht L, Mensink E, de Witte T. In AML t(8;21) colony growth of both leukemic and residual normal progenitors is restricted to the CD34+, lineage-negative fraction. Leukemia 1998; 12: 1782-1788, MEDLINE

65 Miyamoto T, Nagafuji K, Akashi K, Harada M, Kyo T, Akashi T, Takenaka K, Mizuno S, Gondo H, Okamura T, Dohy H, Niho Y. Persistence of multipotent progenitors expressing AML1/ETO transcripts in long-term remission patients with t(8;21) acute myelogenous leukemia. Blood 1996; 87: 4789-4796, MEDLINE

66 Edwards RH, Wasik MA, Finan J, Rodriguez R, Moore J, Kamoun M, Rennert H, Bird J, Nowell PC, Salhany KE. Evidence for early hematopoietic progenitor cell involvement in acute promyelocytic leukemia. Am J Clin Pathol 1999; 112: 819-827, MEDLINE

67 Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, Caceres-Cortes J, Minden M, Paterson B, Caligiuri MA, Dick JE. A cell initiating human acute myeloid leukemia after transplantation into SCID mice. Nature 1994; 367: 645-648, MEDLINE

68 Blair A, Hogge DE, Ailles LE, Lansdorp PM, Sutherland HJ. Lack of expression of Thy-1 (CD90) on acute myeloid leukemia cells with long-term proliferative ability in vitro and in vivo. Blood 1997; 89: 3104-3112, MEDLINE

69 Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nature Med 1997; 3: 730-737, MEDLINE

70 Brendel C, Mohr B, Schimmelpfennig C, Muller J, Bornhauser M, Schmidt M, Ritter M, Ehninger G, Neubauer A. Detection of cytogenetic aberrations both in CD90 (Thy-1)-positive and (Thy-1)-negative stem cell (CD34) subfractions of patients with acute and chronic myeloid leukemias. Leukemia 1999; 13: 1770-1775, MEDLINE

71 Verfaille CM, Miller WJ, Boylan K, McGlave PB. Selection of benign primitive hematopoietic progenitors in chronic myelogeneous leukemia on the basis of HLA-DR antigen expression. Blood 1992; 79: 1003-1010, MEDLINE

72 Jiang X, Lopez A, Holyoake T, Eaves A, Eaves C. Autocrine production and action of IL-3 and granulocyte colony-stimulating factor in chronic myeloid leukemia. Proc Natl Acad Sci USA 1999; 96: 12804-12809, MEDLINE

73 Sokolic RA, Ferguson W, Mark HF. Discordant detection ofmonosomy 7 by GTG-banding and FISH in a patient with Schwachman-Diamond syndrome without evidence of myelodysplasticsyndrome or acute myelogenous leukemia. Cancer Genet Cytogenet 1999; 115: 106-113, MEDLINE

74 Tanaka K, Arif M, Eguchi M, Shintani T, Kumaravel TS, Asaoku H, Kyo T, Dohy H, Kamada N. Interphase fluorescence in situ hybridization overcomes pitfalls of G-banding analysis with special reference to underestimation of chromosomal aberration rates. Cancer Genet Cytogenet 1999; 115: 32-38, MEDLINE

75 Feuring-Buske M, Haase D, Buske C, Hiddemann W, Wormann B. Clonal chromosomal abnormalities in the stem cell compartment of patients with acute myeloid leukemia in morphological complete remission. Leukemia 1999; 13: 386-392, MEDLINE

76 Ritter M, de Kant E, Huhn D, Neubauer A. Detection of DNA methylation in human leukemias by differential polymerase chain reaction. Leukemia 1995; 9: 915-921, MEDLINE

77 Bloomfield C, Shuma C, Regal L, Philip P, Hossfeld D, Hagemeijer A, Garson O, Peterson B, Sakurai M, Alimena G, Berger R, Rowley J, Ruutu T, Mitelman F, Dewald G, Swansbury J. Long-term survival of patients with acute myeloid leukemia. Cancer 1997; 80: 2191-2198, MEDLINE

78 Byrd JC, Dodge RK, Carroll A, Baer MR, Edwards C, Stamberg J, Qumsiyeh M, Moore JO, Mayer RJ, Davey F, Schiffer CA, Bloomfield CD. Patients with t(8;21)(q22;q22) and acute myeloid leukemia have superior failure-free and overall survival when repetitive cycles of high-dose cytarabine are administered (in process citation). J Clin Oncol 1999; 17: 3767-3775, MEDLINE

79 Kwong YL, Ng MH, Ma SK. Familial acute myeloid leukemia with monosomy 7: late onset and involvement of a multipotential progenitor cell (in process citation). Cancer Genet Cytogenet 2000; 116: 170-173, MEDLINE

80 Cignetti A, Bryant E, Allione B, Vitale A, Foa R, Cheever MA. CD34(+) acute myeloid and lymphoid leukemic blasts can be induced to differentiate into dendritic cells. Blood 1999; 94: 2048-2055, MEDLINE

81 Charrad RS, Li Y, Delpech B, Balitrand N, Clay D, Jasmin C,Chomienne C, Smadja-Joffe F. Ligation of the CD44 adhesion molecule reverses blockage of differentiation in human acute myeloid leukemia (see comments). Nature Med 1999; 5: 669-676, Article MEDLINE

82 Saitoh K, Miura I, Takahashi N, Miura AB. Fluorescence in situ hybridization of progenitor cells obtained by fluorescence-activated cell sorting for the detection of cells affected by chromosome abnormality trisomy 8 in patients with myelodysplastic syndromes. Blood 1998; 92: 2886-2892, MEDLINE

83 Janssen JWG, Buschle M, Layton M, Drexler HG, Lyons J, van den Berghe H, Heimpel H, Kubanek B, Kleihauer E, Mufti GJ, Bartram CR. Clonal analysis of myelodysplastic sybdromes. Evidence for multipotent stem cell origin. Blood 1989; 73: 248-254, MEDLINE

Figure 1  Summarization of recent hypotheses of human stem cell development. Model of CD34 expression in the human bone marrow during maturation progress. The earliest stem cell is probably not tissue-determined, quiescent and with self-renewing capacity. Possibly through cell matrix interaction it can acquire a tissue specific phenotype and function. This early hematopoietic progenitor cell with ability for self-renewal seems to be CD34^- CD45⁺ exhibiting variable expression of CD38. On activation this cell can acquire a CD34⁺ phenotype and give rise to differentiated progeny. It displays a strong CD34 expression which is reversible if an activation equilibrium occurs. The activated CD34⁺ cell is partially capable of self-renewal and long-term engraftment, especially if it has a certain pattern of co-expressed antigens indicating an immature developmental stage like CD90⁺, DR^-. As maturation proceeds the co-antigen expression pattern changes and the clonogenic potential improves while self-renewal capacity persists. It should be emphasized that maturation in reality is a continuing process unlike this diagram, which suggests a stepwise maturation with defined stages. Antigen expression is probably a gradually increasing or decreasing process reflecting the fluent maturation progress.