| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Concise Review |
a Department of Microbiology, Immunology, and Molecular Genetics, and
b Howard Hughes Medical Institute, University of California, Los Angeles, California, USA
Key Words. Chronic myeloid leukemia • BCR-ABL • Stem cells
Janusz Kabarowski, Ph.D., HHMI/UCLA, 5-748 MRL Bldg., 675 Charles E. Young Dr. S., Los Angeles, California 90095-1662, USA. Telephone: 310-825-0169; Fax: 310-206-8822; e-mail: januszk{at}microbio.ucla.edu
 |
ABSTRACT |
|---|
 Top Abstract Clinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
|
CLINICAL FEATURES OF CHRONIC MYELOID LEUKEMIA (CML) |
|---|
|
TopAbstractClinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
Although granulocytic expansion is the primary clinical feature in cpCML, megakaryocytic and erythroid progenitor compartments are also expanded; in vitro culture of cpCML bone marrow and peripheral blood reveals elevation of multipotent, erythroid, megakaryocytic and granulocytic clonogenic progenitors. Despite expansion of hematopoietic progenitors, their relative numbers suggest that the lineage potential of cpCML stem cells is not altered [1]. However, the consequences of BCR-ABL expression in the pluripotential hematopoietic stem cell (HSC) and the full extent to which this contributes to the pathogenesis of cpCML are not known. Evaluation is inevitably hampered by the rarity of this cell type and has therefore relied upon surrogate assays of heterogenous populations which include both stem and early multipotential progenitors. An important question is whether the Ph+ stem cell pool is also expanded in cpCML and contributes to the enlargement of subsequent progenitor compartments.
It is assumed that maintenance of the HSC pool by self-renewal can occur by symmetric cell divisions in which each cell gives rise either to two stem cells or two cells committed to differentiation, adoption of alternative fate decisions by the progeny of a symmetric stem cell division, or asymmetric cell division leading to the production of one identical daughter cell and another committed to differentiation. The replicative behavior of single-sorted human fetal-liver derived [2] or human cord-blood derived [3] stem/multipotent progenitor cells suggests that although their mitotic rate can be influenced by extrinsic mechanisms (cytokines), the proportion of asymmetric cell divisions during early hematopoiesis may be primarily intrinsically determined, at least in vitro. Although both stochastic and instructive mechanisms for lineage commitment of HSCs have been proposed, most studies support the former with respect to the initial "activation" event, and do not exclude an instructive role for growth factors and other external influences downstream [4-6]. The signals, both intrinsic and extrinsic, controlling these and other fate decisions in stem cells are largely unknown, but may be influenced by one or more biological processes upon which BCR-ABL tyrosine kinase activity impinges.
|
HEMATOPOIETIC STEM AND PROGENITOR POPULATIONS IN CPCML |
|---|
|
TopAbstractClinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
Early studies performed on B lymphoblastoid lines derived from cpCML blood demonstrated that most B cell lines expressing allelic biochemical markers in common with the leukemic clone harbored multiple chromosomal abnormalities, but did not show evidence of the Ph translocation by cytogenetic examination [11]. This suggested that clonal expansion may occur prior to the acquisition of the Ph chromosome in some patients, which raises the possibility that BCR-ABL expression may not be the initiating or sole event in the development of CML. More recent studies in which BCR-ABL transcripts have been detected by optimized polymerase chain reaction analysis in very small numbers of peripheral blood leukocytes from 30%-70% of normal healthy adults raise similar questions [12, 13]. However, it is possible that this aberrant recombination event occurs in mature precursors and does not precipitate the onset of leukemia, as their progeny are eventually eliminated through differentiation and apoptosis. Further studies are required to establish the phenotype of these cells and whether this phenomenon is also observed in more primitive cells in the bone marrow of normal subjects.
Despite these interesting observations, it is widely accepted that acquisition of the Ph translocation and expression of the BCR-ABL tyrosine kinase is the initiating and rate-limiting event in CML. This view is supported by many observations made in studies of BCR-ABL expression in hematopoietic and fibroblastic cell lines as well as animal model systems. Also, certain clinical features of cpCML patients have been interpreted as inconsistent with the notion that acquisition of the Ph chromosome is a secondary aetiological event in the pathogenesis of CML. For example, recent studies employing fluorescence-activated cell sorting of CD34+ HLA-DR– stem and progenitor cell fractions from the bone marrow and peripheral blood of cpCML patients have failed to detect clonal Ph– hematopoiesis [14]. In addition, polyclonal hematopoiesis is restored following interferon-
(IFN-)-induced cytogenetic remission. However, one cannot exclude the possibility that this, as well as other treatment regimens, may be acting upon events other than, and anteceding, BCR-ABL expression. Nevertheless, the efficacy with which the ABL-specific tyrosine kinase inhibitor STI 571 induces hematological and cytogenetic remission in patients also strongly supports the view that BCR-ABL expression is the rate-limiting event in CML [15]. However, it should be stressed that long-term follow-up of STI 571-treated patients is required to determine whether any develop Ph– leukemias.
An important feature of cpCML is the coexistence of normal hematopoiesis [16, 17]. Although this is competed for and progressively replaced by the leukemic clone, patients in early chronic phase often have detectable Ph– stem and progenitor cells in their bone marrow, and normal hematopoiesis can be temporarily restored following chemotherapy, IFN- treatment, or after long-term culture ex vivo [14, 16, 18]. Much effort is being made to optimize protocols for purging leukemic bone marrow of Ph+ cells and enrichment of normal stem cells prior to autologous transplantation. Such procedures must inevitably exploit differential biological responses of Ph+ versus Ph– stem cells to ex vivo manipulation and exogenously applied factors such as early-acting cytokines, growth inhibitory cytokines, and culture conditions. While promising results have thus far been achieved in terms of periods of disease-free survival, the underlying mechanisms are not understood, despite many candidate biological and signal transduction pathways being linked to BCR-ABL expression.
Are Ph+ Stem Cell Numbers Abnormally High in cpCML?
A major proportion of the myeloid expansion in cpCML can be attributed to enlargement of the colony-forming units-granulocyte, macrophage compartment [19, 20]. However, increased cell numbers within a more primitive population which includes cells with stem properties in long-term culture, are also present in cpCML, although they exhibit a reduced probability of differentiation in vitro [21]. It is possible, therefore, that increased stem cell self-renewal in vivo at the expense of differentiation also contributes to the cellular expansion, a scenario in which expansion of the stem cell population eventually compensates for the reduced probability of differentiation.
Protocols for the enrichment of HSCs based on surface expression of antigens have proven valuable in furthering the biological characterization of stem and progenitor compartments [22]. However, the populations obtained exhibit considerable heterogeneity with respect to clonogenic potential in long-term culture and repopulating ability in immunodeficient mice. Indeed, quantitative analysis of long term culture-initiating cells (LTC-IC) and high proliferative potential-colony-forming cells in these populations only provides a correlative measurement of stem cell numbers. In addition, recent evidence demonstrating that cell-surface expression of CD34 is reversible and does not define all stem cells capable of long-term engraftment [23-25] suggests that use of this marker in positive selection protocols may exclude a subpopulation of cells with true stem cell activity [26]. Nevertheless, studies have shown that Ph+ cells are only rarely detected within bone marrow CD34+, HLA-DR– fractions obtained from cpCML patients, which are generally Ph– and polyclonal [14, 27, 28]. Indeed, the frequency of Ph+ LTC-ICs in the bone marrow of most cpCML patients is lower than that of their Ph– counterparts in normal individuals [29]. However, this Ph– primitive population diminishes with disease progression and Ph+ monoclonal cells become more apparent, consistent with the progressive outcompetition of normal hematopoiesis by the leukemic clone in cpCML. Importantly, although Ph+ LTC-ICs retain normal in vitro clonogenic differentiative potential, they have a marked growth/survival disadvantage in culture [29-31], making estimation of Ph+ stem cell frequency in patient bone marrow by this surrogate assay difficult and most likely an underestimate.
The frequency of Ph+ hematopoietic repopulation in serially transplanted immunodeficient mice may be a better corollary of stem cell frequency. However, engraftment of cpCML cells in severe combined immunodeficient/ nonobese diabetic (SCID/NOD) mice has proven difficult, with predominance of Ph– hematopoiesis in reconstituted recipients [32, 33]. Recently refined methods have led to better long-term reconstitution [34]. Ph+ stem cells may engraft poorly due to an intrinsic defect in their homing and migratory responses. Although BCR-ABL expression in cell lines increases their spontaneous cellular migration, it actually inhibits chemotactic responses to stromal derived factor-1 (SDF-1) [35]. The receptor for SDF-1, CXCR4, is essential for the appropriate localization of stem and progenitor cells within the bone marrow and efficient engraftment of SCID repopulating stem cells [36, 37]. While perhaps contributing to the premature release and abnormal trafficking of Ph+ progenitors in cpCML, impaired SDF-1-induced chemotaxis also suggests that the impact of BCR-ABL expression upon stem cell frequency may not be adequately represented by analysis of long-term repopulation in transplanted recipient mice.
A major obstacle in establishing any abnormal features of cpCML stem cells more definitively has been our limited understanding of the regulatory factors and interactions which control normal stem cell behavior. However, perturbations in normal proliferative, survival and differentiative control have been widely reported in cultured Ph+ hematopoietic progenitors under a variety of conditions in which the biology of their normal counterparts is much better characterized [19]. Perhaps these defects can therefore be viewed more confidently as in vitro corollaries of pathological features in cpCML and, importantly, suggests impact of BCR-ABL upon certain signaling pathways, many of which have been implicated in the leukemogenic process by studies utilizing in vitro transformation and animal model systems.
|
IN VITRO TRANSFORMATION AND ANIMAL MODEL SYSTEMS |
|---|
|
TopAbstractClinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
These interactions may underlie the decreased adhesion of cpCML progenitors to bone marrow stromal cells [47, 48]. Stromal contact has been shown to reduce the probability of quiescent progenitor cells entering cell cycle and limit the expansion of LTC-ICs [49, 50] and so this may contribute to enlargement of the progenitor cell pool in cpCML. Indeed, CD34+ cpCML cells continue to proliferate in contact with stroma [51], and restoration of integrin-dependent adhesion by treatment with an activating anti-ß1 integrin antibody also restores stromal inhibition of proliferation [52]. Abnormal stromal interaction may also underlie the demise of Ph+ LTC-ICs in culture. The hypercellularity of the bone marrow in cpCML probably contributes to the suppression of normal stromal-dependent Ph– hematopoiesis. Yet an important question is whether Ph+ stem cells also exhibit altered stromal interactions which could deregulate homeostatic mechanisms governing their numbers.
Proliferative Signals
Expression of BCR-ABL in hematopoietic cell lines abrogates their growth factor requirement for survival and proliferation [53-55]. Although cpCML progenitors are generally not growth-factor independent or hyperresponsive to cytokines [56], subtle alterations in their response to combinations of growth factors, including stem cell factor (SCF), erythropoietin (Epo) and GM-CSF, as well as differentiative anomalies have been reported [57, 58]. These observations suggest that BCR-ABL expression in primary hematopoietic cells may activate signaling pathways that elicit lineage and differentiation stage-specific responses.
Activation of RAS, PI 3' kinase, STAT, ERK, and JNK mitogen activated protein (MAP) kinase pathways is observed in BCR-ABL-transformed cell lines representative of myeloid, lymphoid, and multipotential progenitor cells. Although controversy exists regarding the ascription of certain signaling molecules as downstream effectors of BCR-ABL, activity of the RAS/RAF, PI 3' kinase/Akt, and JNK pathways has been suggested to be required for efficient transformation and leukemogenesis induced by BCR-ABL [59-61]. However, JNK and not ERK is activated by induction of a temperature-sensitive BCR-ABL mutant in a hematopoietic cell line or following transient coexpression with BCR-ABL in COS or 293T [62, 63]. Inducible expression of BCR-ABL in fibroblasts, on the other hand, is associated with a concomitant increase in ERK activity [64], suggesting that BCR-ABL may not directly activate the RAS/ERK MAP kinase pathway in certain cell types. Tyrosine kinase-dependent activation of RAS by the cytoplasmically localized BCR-ABL protein may require recruitment of a sufficient portion to the plasma membrane within integrin-dependent focal adhesion complexes which are less developed in nonadherent hematopoietic cell lines. Nevertheless, multiple potential routes from BCR-ABL to RAS have been defined based upon its interaction with upstream adaptor molecules and the RAS dependence for activation of several components of MAP kinase modules by BCR-ABL in transient expression systems [65, 66]. In addition, BCR-ABL might activate RAS-dependent pathways through inhibition of p120 GAP via p62 DOK, a major substrate of BCR-ABL in cpCML progenitors [67].
While there is evidence both for and against the utilization of several key signaling pathways by BCR-ABL, their evaluation in primary cells has proven difficult. Approaches undertaken include coexpression of dominant negative mutants of candidate signaling molecules with BCR-ABL in murine bone marrow transformation and leukemogenesis systems [46, 60, 68]. However, studies using dominant negative mutants to examine the requirement for cytoplasmic signaling molecules have rarely, if ever, addressed the issue of BCR-ABL specificity by assessing their impact upon transformation of the same hematopoietic cell type by an unrelated nuclear oncogene. Therefore, although a requirement for a particular signaling protein can be established, its assignment as a direct downstream effector of BCR-ABL cannot be made without more direct evidence of its activation.
Activation of pathways associated with mitogenic responses suggests that BCR-ABL may provide a direct proliferative stimulus. Although a few studies have demonstrated such an effect [64, 69], the features of cpCML do not support this view. For example, although Ph+ progenitors in cpCML generally proliferate continuously, in vivo cell kinetic studies do not indicate that cpCML progenitors proliferate more rapidly than normal [70, 71]. In fact, several studies suggest that cpCML myeloid and erythroid progenitors traverse cell cycle more slowly [72, 73], although they undergo extra rounds of cell division prior to terminal differentiation according to the "discordant maturation" hypothesis of Clarkson and Strife [74]. Whether discordant nuclear and cytoplasmic maturation is a feature of differentiating Ph+ stem cells is difficult to address directly. However, it has recently been suggested, based upon analysis of telomere length as a surrogate marker of stem cell replicative history in Ph+ versus Ph– cells from the same individual, that they undergo a greater number of cell divisions than their Ph– counterparts [75].
Importantly, quiescent (G0) Ph+ stem cells are detectable in cpCML [76], suggesting that the presence of BCR-ABL does not obligingly promote their entry into cell cycle and normal responses to proliferative and differentiative stimuli may be retained by Ph+ stem cells to a significant degree. This, of course, is a major consideration in the design of successful transplantation protocols and purging of Ph+ cells based upon proliferative indices [30]. In fact, sustained activation of mitogenic signals might be expected to deplete the HSC pool by driving cells from quiescence into cell cycle. This is supported by the recent observation that the deregulated proliferation of HSCs in p21cip1/waf1 -/- mice results in their depletion under conditions of stress and following transplantation [77].
The level and duration of signaling are critical determinants of biological response and one cannot rule out the possibility that the developmental stage of HSCs imparts a unique responsiveness to these signals. Additionally, the level of expression of BCR-ABL may significantly influence the cellular response (e.g., proliferation versus suppression of apoptosis) [78, 79]. In this regard, it is noteworthy that duplication of the Ph chromosome is a common feature of disease progression in CML [80].
Altered responsiveness of stem and progenitor populations to growth-promoting or growth-inhibitory factors may also contribute to the cellular expansion in cpCML. Rather than providing a direct mitogenic stimulus, BCR-ABL may influence proliferative and survival control mechanisms by modulating upstream cytokine and growth factor receptors. For example, by direct receptor phosphorylation or subcellular redistribution of key signal transduction molecules, BCR-ABL may alter the duration and/or threshold of activation of proliferative signaling pathways such as RAS/ERK MAP kinase, which are normally transiently activated to a lower level in a growth factor or adhesion-dependent manner. BCR-ABL physically associates with the SCF receptor, c-Kit [81], and cpCML progenitors exhibit a similar spectrum of tyrosine-phosphorylated proteins to that elicited in normal progenitors in response to SCF, including c-Kit itself [82]. Unlike their normal counterparts, the primitive CD38– fraction of CD34+ cells from cpCML patients can be induced to proliferate in SCF in the absence of supplementary cytokines [83]. It has been suggested that this abnormal responsiveness to SCF may be due to the constitutive activation of one or more downstream cytokine receptor signaling pathways which normally synergize with SCF/c-Kit-mediated signals by BCR-ABL, rather than the amplification of, or increased sensitivity to, c-Kit-dependent signaling [84]. Although autocrine cytokine production has been demonstrated in BCR-ABL-transduced bone marrow cells and short-term cultures of CD34+ cpCML cells [69, 85], secretion of cytokines by cpCML cells does not underlie this phenomenon [83].
In addition, despite the tremendous promise of STI 571 as a treatment for CML patients [15], this kinase inhibitor also inhibits c-Kit receptor tyrosine kinase activity [86], which raises the possibility that although the combination of BCR-ABL and c-Kit inhibition may contribute to the efficacy of this compound in CML, its long-term impact upon Ph– stem cells and progenitors may warrant evaluation.
Antiapoptotic Signals
Despite reports of normal apoptotic responses of cpCML progenitors to growth factor deprivation and genotoxic treatment [87, 88], it is widely held that suppression of apoptosis is a feature of cpCML progenitors [89]. Many studies have demonstrated an antiapoptotic function for BCR-ABL in hematopoietic cell lines and murine bone marrow cells. Importantly, in hematopoietic cell lines this property is separable from their transformation to growth factor independence [55, 62], suggesting a direct impact of BCR-ABL tyrosine kinase activity upon survival pathways. Signals implicated in the antiapoptotic function of BCR-ABL include PI 3' kinase/Akt [60, 90], STAT5-dependent BCLXL upregulation [91] and NF-
B [92]. In addition, prolonged activation of the G2-specific cell cycle checkpoint via maintenance of inhibitory phosphorylation of CDC2 at tyrosine 15 has been ascribed directly to BCR-ABL, and shown to underlie its ability to suppress apoptosis and promote the long-term survival of hematopoietic cells following DNA damage [93, 94].
Although the impact of BCR-ABL-derived antiapoptotic signals at the level of the HSC is not known, recent studies with transgenic mice expressing BCL2 under the control of the H-2Kb promoter provide an interesting insight into the consequences of a strong sustained antiapoptotic signal in this cell type. Enforced expression of BCL2 promotes the gradual expansion of HSCs despite inhibiting their entry into cell cycle [95]. This suggests that mechanisms other than direct stimulation of G0/G1
S transition can influence HSC frequency.
Animal Model Systems
Taken together, it seems most likely that the expansion of progenitors observed in cpCML results from the impact of BCR-ABL upon more than one biological process, each intimately interrelated, making it conceivable that perturbation of any one by BCR-ABL is sufficient to dysregulate one or more of the others. This scenario is consistent with the multiple, often subtle, defects observed in different cell types in cpCML, including dysplastic abnormalities and increased life span in granulocytes [96], adhesive defects in progenitors and altered differentiative responsiveness of erythroid precursors to Epo [19]. Their consequences in the pluripotential HSC are largely uncharacterized, due primarily to the lack of a suitable animal model system.
A variety of transgenic models expressing BCR-ABL from global and lineage-restricted promoters develop leukemias with varying latency and penetrance [97-99]. Introduction of BCR-ABL into murine bone marrow populations by retroviral transduction followed by transfer into irradiated hosts gives rise to a spectrum of hematological malignancies including granulocytic hyperplasias resembling cpCML [100, 101]. While differences in genetic background and retroviral infection protocols probably underlie the variability observed in the type, latency, and transplantability of leukemias developing in early animal models, a failure to efficiently target BCR-ABL expression to HSCs is probably the major factor. Recently improved methods for retroviral transduction of bone marrow populations enriched for stem cells [102] and the development of transgenic mice in which BCR-ABL expression is under the control of lineage-specific promoters [98] have provided more reproducible models of cpCML-like myeloproliferative disease. The current availability and future development of many gene-targeted mouse strains validate a genetic approach to evaluate mechanisms by which BCR-ABL deregulates normal hematopoietic control [103-105].
Nevertheless, targeting expression of BCR-ABL via HSC-specific promoter or enhancer elements in transgenic mice still falls short of truly recapitulating the initiating event in the pathogenesis of CML, that is the acquisition of the Ph translocation and BCR-ABL expression in an adult pluripotential HSC later in life. Indeed, certain features of HSCs in old mice are distinguishable from those in young mice, most notably increased cycling [106]. Extrapolating to the human disease, this may be a factor which actually underlies the age-related incidence of CML. Controlled induction of BCR-ABL expression in HSCs of the adult mouse is therefore required. Although a recent tetracycline-regulatable transgenic system demonstrates that BCR-ABL alone is sufficient to induce leukemia and its continued expression is required for maintenance of the leukemic phenotype, BCR-ABL expression is induced via a mouse mammary tumor virus long terminal repeat-linked tetracycline transactivator, resulting in the rapid and exclusive development of Pro-B cell leukemias [107]. Nevertheless, this approach holds much promise and tetracycline-regulatable transgenic expression of BCR-ABL in HSCs via promoter or enhancer elements derived from HSC-specific genes such as scl will not only more accurately model the development of CML, but will also provide a system with which to evaluate the impact of BCR-ABL upon stem cells in a dose- and time-dependent manner.
A tetracycline-regulatable expression system in embryonic stem (ES) cells and their hematopoietic progeny recently developed in our laboratory [108] provides a readily manipulatable system with which to study these and other important issues. This system clearly demonstrates expansion of ES cell-derived stem/multipotent (c-Kit+, Lin–, and CD34+, Lin–) and myeloid progenitors with concomitant suppression of erythroid colony development in differentiating stromal (OP9)-dependent cultures induced to transiently express BCR-ABL, consistent with clinical features of the human disease. Perhaps the most attractive feature of this model is the ease with which large numbers of candidate primitive stem/progenitor cells can be derived and subsequently induced to express varying doses of BCR-ABL in a completely reversible manner. Current efforts are therefore directed towards assessing the direct impact of regulated BCR-ABL expression upon proliferative and differentiative control mechanisms in these candidate pluripotential HSCs by assaying their repopulating potential in transplanted recipients subjected to regulated tetracycline administration and withdrawal. It is hoped that this and similar experimental approaches will provide insights into hitherto unanswered questions regarding the biology of the HSC in CML, and the impact of biological phenomena established from studies of BCR-ABL in heterologous cell types at this level of hematopoietic development. A better understanding of these issues may uncover novel mechanisms underlying the efficacy of therapeutic agents such as STI 571 and also lead to improved regimens for purging malignant cells from autografts. We hope the aetiological target in CML will prove to be a less elusive therapeutic target in the future.
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractClinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
|
References |
|---|
|
TopAbstractClinical Features of Chronic...Hematopoietic Stem and...In Vitro Transformation and...References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
