If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Departments of Haematology and Transplant Services, Australian Redcross Blood Service South Australia and Division of Haematology, IMVS, Adelaide, South Australia
Chronic myelomonocytic leukemia (CMML) is a heterogeneous disease with no effective treatments or cure. Several factors have been implicated in its pathogenesis. In the current study, we studied the dependence of CMML on granulocyte-macrophage colony-stimulating factor (GM-CSF).
Materials and Methods
We used in vitro colony assays in methylcellulose where CMML cells were tested in the presence or absence of the specific GM-CSF antagonist E21R. We also developed an in vivo model in which CMML cells were tested for their ability to engraft into immunodeficient mice transgenic for human GM-CSF.
Results
Bone marrow cells from seven of seven patients with CMML formed spontaneous colonies that were sensitive to E21R treatment, with reduction in colony growth by up to 92%. E21R also inhibited colony formation by CMML patient cells stimulated by exogenously added GM-CSF but not interleukin-3. In in vivo experiments we observed engraftment of CMML cells (but not normal cells) in immunodeficient mice transgenic for human GM-CSF. None engrafted in nontransgenic mice. Cell dose escalation showed that the optimal number was 0.5 to 1 × 108 peripheral blood mononuclear cells per mouse, which is equivalent to an infusion of 0.2 to 3.6 × 106 CD34+ cells. Time course experiments showed that maximal engraftment occurred 6 weeks after injection.
Conclusions
These results demonstrate that in some CMML patients, GM-CSF produced by either autocrine or paracrine mechanisms is a major growth determinant. The results suggest that therapies directed at blocking this cytokine could control the growth of some CMML patients in vivo.
Chronic myelomonocytic leukemia (CMML) represents a distinct myelodysplastic syndrome (MDS) in which dysplasia is observed in the bone marrow (BM), and an excess of monocytes is present in both the blood and BM. The mechanism for the monocytosis remains unknown, and the disease itself is highly heterogeneous. Most patients exhibit lethargy, infection, or exaggerated bleeding, and up to 25% evolve into acute myeloid leukemia (AML) at 3 to 5 years. For the remainder of patients, median survival is between 5 and 24 months [
]. The World Health Organization recently included CMML within the new category of MDS/myeloproliferative disorder (MPD), along with juvenile myelomonocytic leukemia (JMML) and atypical chronic myeloid leukemia (CML) [
World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues report of the Clinical Advisory Committee meeting-Airlie House, Virginia, November 1997.
] in the myeloid neoplasms classification. Current treatment options for CMML are highly unsatisfactory; the only curative option is allogeneic BM transplantation. Given the elderly nature of the patient population, this is rarely a serious clinical option [
], whereas a number of clinical studies have addressed the potential for cytokines to improve hematopoietic maturation. Results of trials with granulocyte-macrophage colony-stimulating factor (GM-CSF) have been disappointing, showing increased rates of transformation to AML [
]. For treatment of anemia associated with CMML, human erythropoietin (Epo) has been administered, but with response in only 15 to 20% of patients (detailed in [
], no single etiologic agent has been linked to the development of CMML. Mutations in Ras are seen more often in those patients with CMML than any other subtype of MDS [
], but they are not universal. Also, a variety of cytogenetic defects have been noted. For example, TEL, a member of the Ets family of transcription factors, has several fusion partners in hematologic malignancies (reviewed in [
Transformation of hematopoietic cell lines to growth-factor independence and induction of a fatal myelo- and lymphoproliferative disease in mice by retrovirally transduced TEL/JAK2 fusion genes.
] or platelet-derived growth factor receptor β (PDGFβR), can lead to a myeloproliferative disease in mice. The t(5;12)(q33;p13) translocation generates the TEL/PDGFβR fusion protein [
The TEL/platelet-derived growth factor b receptor (PDGFβR) fusion in chronic myelomonocytic leukemia is a transforming protein that self-associates and activates PDGFbR kinase-dependent signaling pathways.
Interleukin-10 inhibits spontaneous colony-forming unit growth from human peripheral blood mononuclear cells by suppression of endogenous granulocyte-macrophage colony-stimulating factor release.
Interleukin-10 inhibits spontaneous colony-forming unit growth from human peripheral blood mononuclear cells by suppression of endogenous granulocyte-macrophage colony-stimulating factor release.
] have been implicated given their stimulation of CMML growth in vitro and their presence in the serum of some CMML patients.
A second limitation has been the absence of an animal model with which to study engraftment and progression of CMML cells in vivo. Transplantation of human hemopoietic cells into immunodeficient mice, such as the nonobese diabetic severe combined immunodeficient (NOD/LtSz-scid/scid [NOD/SCID]) mouse [
Inhibition of granulocyte-macrophage colony-stimulating factor prevents dissemination and induces remission of juvenile myelomonocytic leukemia in engrafted immunodeficient mice.
], but no data on requirements for CMML engraftment have been reported. By generating a colony of NOD/SCID mice transgenic for human GM-CSF, we show that engraftment of CMML cells is improved by the presence of GM-CSF, and the conditions for optimal engraftment are described. These results, in addition to the in vitro study, suggest that GM-CSF is a common pathogenic determinant in the development of CMML in patients who do not exhibit constitutively activating chromosomal abnormalities and that its antagonism [
] may represent a general and effective therapeutic option for these patients.
Materials and methods
Patient samples
These studies were performed using BM and peripheral blood (PB) cells from normal donors or patients with CMML after obtaining informed consent (approved by the Royal Adelaide Human Ethics Committee). From 10 to 25 mL of BM was aspirated from the posterior superior iliac crest and up to 100 mL of PB was collected in preservative-free heparin. At the time the patient was enrolled in the study, a full clinical history was recorded. For normal patient samples, BM was aspirated into preservative-free heparin from the sternum and posterior iliac crest of normal healthy volunteers.
Cytokines and monoclonal antibodies
Human GM-CSF and human IL-3 were produced in the laboratory as previously described [
Barry SC, Bagley CJ, Phillips J, et al. (1994) Two contiguous residues in human interleukin-3, Asp21 and Glu22, selectively interact with the α- and β-chains of its receptor and participate in function. J
]. The antibodies used in this study were directed against the GM-CSF receptor βc chain (1C1) or against α chains of the GM-CSF (4H1) and IL-3 receptors (7G3). We also used the recombinant GM-CSF analogue E21R [
] (produced by BresaGen Ltd., Thebarton, Australia).
Colony assays
BM mononuclear cells from CMML patients or normal BM donors were purified over a Ficoll-Hypaque gradient (1.077 g/mL, Lymphoprep; Nycomed, Oslo, Norway) and washed three times with RPMI (JRH Biosciences, Victoria, Australia) supplemented with 10% heat-inactivated fetal calf serum [HI-FCS]); Trace Biosciences Pty Ltd., Castle Hill, NSW, Australia). Cells were cryopreserved in 10% dimethylsulfoxide and stored in liquid nitrogen. Samples were rapidly thawed into IMDM (JRH Biosciences) containing 20% HI-FCS and DNAse at 500 U/mL (Sigma, Castle Hill, NSW, Australia) and washed twice in the same medium. Colony assays were performed culturing 100,000 BM cells in 0.8% methylcellulose, 30% FCS, 1% bovine serum albumin (BSA), 10−4 M 2-mercaptoethanol, and 2 mM glutamine (Methocult; Stem Cell Technologies, Vancouver, Canada) supplemented with cytokines as shown on figures. Plates were incubated for 14 days at 37°C in an atmosphere of 5% CO2 in air. Colonies were enumerated in the same dish using an inverted microscope.
Lineage analysis of BM colonies
For each dish, 100,000 BM cells were plated into a top agar layer (0.5 mL) over a bottom feeder agar layer containing cytokines (1.1 mL). The cytokines were identical to those outlined for methylcellulose cultures. Plates were incubated for 14 days at 37°C in an atmosphere of 5% CO2 in air and then fixed overnight at 4°C using 1.5% glutaraldehyde. Agar plugs were transferred to glass slides and stained with Luxol Fast Blue MBS (BDH, Poole, England), Fast Garnet GBC (Sigma), and Fast Blue BB (Sigma). Slides were dried and mounted in CV Mount (Leica, Heidelberg, Germany).
Immunodeficient mice
NOD/SCID mice were obtained from the Walter and Eliza Hall Institute (Melbourne, Australia) and maintained under sterile conditions at the University of Adelaide Laboratory Animal Services. All mice were housed in microisolator cages in HEPA-air filtered rooms in specific pathogen-free conditions and provided with autoclaved food and water ad libitum. Experiments were performed as approved by the Animal Ethics Committee of the University of Adelaide.
Generation of mice transgenic for human GM-CSF
Plasmid constructs containing the GM-CSF cDNA under the control of the TCR enhancer/H2-Kb promoter were produced by BresaGen. The constructs were isolated from vector sequences and coinjected into the male pronucleus of fertilized mouse eggs by standard techniques [
Hogan B, Beddington R, Constantini F, Lacy E (1994) Manipulating the Mouse Embryo. A Laboratory Manual. New York: Cold Spring Harbor Laboratory Press, p. 217
]. Transgenic founder mice were identified by DNA Southern blotting of human GM-CSF gene from DNA isolated from tail samples of the mice. Transgenic C57BL/6 founders were crossed with homozygous NOD/SCID mice, and the progeny carrying the transgenes were back-crossed with homozygous NOD/SCID mice for seven or more generations. Expression of the transgene messenger RNA was demonstrated by reverse transcriptase polymerase chain reaction (RT-PCR) from BM, spleen, liver, kidney, brain, heart, PB, and lung as described [
]. From 1 to 10 μg of total RNA was reverse transcribed with 300 units of Superscript II, 1 × first strand buffer (Gibco-BRL, Melbourne, Australia), 10 mM DTT, 500 mM dNTPs, and 500 ng of (dT)17-ADAPTOR primer [
] at 42°C for 1 hour. the 3′-RACE reactions contained 1 unit AmpliTaq (Applied Biosystems, Victoria, Australia), 1 × Taq buffer, MgCl2 to a final concentration of 1.5 mM, 200 μM dNTPs, 100 ng of the human GM-CSF specific primer (5′-TAGAGACACTGCTGCTGAG-3′), 100 ng of the ADAPTOR primer (5′-GACTCGAGTCGACATCG-3′), and 2 μL of first-strand cDNA template. Amplification was achieved by incubation at 94°C for 2 minutes, followed by 35 cycles of 94°C (30 seconds), 55°C (30 seconds), and 72°C (2 minutes). Additionally, mice were confirmed to be producing human GM-CSF using an enzyme-linked immunosorbent assay kit (R & D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions. The mice used in this study had human GM-CSF levels of 100 ng/mL in their serum, which was maintained with the successive back-crosses to NOD/SCID.
Transplantation of human cells into immunodeficient mice
Eight-week-old NOD/SCID mice were irradiated with a sublethal dose of total body gamma irradiation (300–350 cGy) using a deep x-ray irradiation machine with a dose rate of 100 cGy/min. Mononuclear cells obtained from the PB from CMML patients or normal subjects were injected into the irradiated mice via tail vein. Cells from five patients were split into two aliquots and injected into transgenic mice and into normal mice. Samples from different patients were not mixed. Mice were housed with daily observation to identify the onset of disease (weight loss, anemia, or cachexia).
Providing they remained healthy, mice were sacrificed 4 to 10 weeks after cellular infusion and engraftment was analyzed by flow cytometry. BM and spleen cells were collected in RPMI 1640 supplemented with 2 to 10% FCS. Single-cell suspensions were assessed for the presence and phenotype of human cells by immunolabeling with conjugated antibodies to human cell surface proteins. Antibodies used for analysis of each animal included fluorescein isothiocyanage (FITC)-conjugated anti-IgG/phycoerythrin (PE)-conjugated anti-IgG (negative control provided with the LeucoGATE), FITC-conjugated anti-CD45/PE-conjugated anti-CD14 (LeucoGATE), FITC-conjugated anti-CD34 (anti-HPCA-2), PE-conjugated anti-CD33, FITC-conjugated anti-CD3 (all from Becton-Dickinson), and RPE-conjugated anti-CD19, Clone HD37 (Dako). Cells were incubated with antibodies for 45 to 60 minutes in the dark and on ice and then were washed with RPMI/5% FCS before being fixed with FACS Fix (1% formaldehyde, 2% D-glucose, 0.02% sodium azide in phosphate-buffered saline). All samples were analyzed using a Coulter EPICS XL-MCL (Beckman Coulter, NSW, Australia) and gated to include both the blast/lymphocyte populations and the monocyte cells. Quantitation of human cells for all engrafted mice was compared with both control animals that had received no transplant of human cells, and to the individual animal's own cells stained with negative control antibodies as outlined earlier.
Results
Patient characteristics
We tested seven patients for in vitro growth of BM cells. Table 1 summarizes the main laboratory findings of these patients. Peripheral blood from two of the original seven along with 15 other patients with CMML was used for in vivo studies. One characteristic of CMML is the male predominance of patients, which is reflected by our data. We received 70% (14/20) of our samples from males with an overall average age of 70.8 years (range 48–87 years). Cryopreserved BM from the seven patients and PB from all 15 patients with CMML was analyzed by flow cytometry to assess CD34 levels. Levels ranged from 1.4 to 11.7% (n = 7; average 4.7%) for BM samples and from 0.2 to 19.8% for PB cells (n = 15; average 3.9%). CD34 expression on the normal BM samples included in this study ranged from 2.4 to 16.1% (n = 8; mean average 9.5%) and normal PB samples ranged from 0.1 to 0.3% (n = 4; average 0.2%). Expression of the GM-CSF receptor α and β subunits was verified on all normal and CMML samples using monoclonal antibodies generated against the two subunits of the GM-CSF receptor as previously described [
Monoclonal antibody 7G3 recognizes the N-terminal domain of the human interleukin-3 (IL-3) receptor α-chain and functions as a specific IL-3 receptor antagonist.
CMML patient samples exhibit spontaneous colony formation that is inhibited by E21R
To assess the clonal potential of BM cells from patients with a proven diagnosis of CMML, samples were analyzed in a semisolid methylcellulose assay. Culturing cells from CMML patients in the absence of exogenous cytokines led to spontaneous formation of colonies from all patients tested, with numbers ranging from 8 to 62 colonies per 100,000 BM cells plated (Fig. 1). Morphologic examination in two patients revealed that these were composed of 53.1 ± 5.2% macrophage colonies and 40.0% ± 5.1% neutrophil colonies. Addition of the specific GM-CSF antagonist E21R reduced spontaneous colony formation in all CMML patient samples tested. Using E21R at 1μg/mL, the number of colonies was reduced by 14 to 79%, whereas reduction using E21R at 10 μg/mL ranged from 49 to 91% (Fig. 1). Using Wilcoxon's matched pair analysis, the reduction in colony number was statistically significant (p < 0.05). As a control we plated BM cells from normal donors that, under identical conditions, showed very few or no spontaneous colonies (range 0–3.7 [average 1.4] colonies per 100,000 cells plated).
Figure 1Spontaneous colony formation by CMML bone marrow cells is inhibited by the GM-CSF antagonist E21R. A total of 100,000 bone marrow mononuclear cells from CMML patients were plated in 35-mm dishes in 1-mL methylcellulose in the absence of any factors or in the presence of E21R at 1 or 10 μg/mL. Data show seven individual patients with triplicate plates ± 1 SEM.
E21R inhibits GM-CSF– but not IL-3–stimulated CMML cells
To examine whether CMML colony growth could be stimulated further by exogenous cytokines and as a specificity control for E21R, we performed colony assays in the presence of GM-CSF or IL-3. Increasing amounts of GM-CSF stimulated further colony growth from some of the patient samples. This was statistically significant (p < 0.05) in patients 2, 4, 1, and 6 but not in patients 3, 5, and 7, suggesting that they had already reached their maximum colony formation potential in the absence of exogenous GM-CSF (Fig. 2). E21R consistently reduced the number of BM colonies with levels of inhibition progressively declining as the concentration of added GM-CSF increased. E21R (10 μg/mL) reduced colony number significantly with GM-CSF added at 0.1 ng/mL (p < 0.05), but this was overcome when the concentration of GM-CSF was increased to 1 or 10 ng/mL. Colony levels were not significantly reduced by E21R with GM-CSF at either 1 or 10 ng/mL.
Figure 2Further stimulation of CMML colony growth by GM-CSF and inhibition by E21R. A total of 100,000 bone marrow mononuclear cells from CMML patients were plated in 35-mm dishes in 1-mL methylcellulose in the absence of any factors or in the presence of GM-CSF, which was added to dishes at the concentration shown. E21R was added to appropriate dishes at 10 μg/mL.
IL-3 also stimulated CMML colony formation with E21R, only marginally reducing colony numbers (Fig. 3). Because the antagonism of E21R is specific for GM-CSF, its colony reduction may be due to a direct apoptotic effect [
] or to inhibition of the spontaneous colonies formed by CMML cells. The reduction in colony number was not significant for either of the two levels of IL-3 tested. In control experiments using BM cells from normal donors, we found that E21R inhibited colony formation in cells stimulated by GM-CSF but not by IL-3 (data not shown).
Figure 3Stimulation of CMML colony growth by IL-3. A total of 100,000 bone marrow mononuclear cells from CMML patients were plated in 35-mm dishes in 1-mL methylcellulose in the absence of any factors or in the presence of IL-3. IL-3 was added to dishes at the concentration shown, in the absence or presence of E21R at 10 μg/mL.
Having shown that these CMML cells were dependent on GM-CSF for growth in vitro, we then investigated whether GM-CSF was necessary in vivo. We developed transgenic mice carrying the gene for human GM-CSF under the control of the mouse H-2Kb promoter. These mice were generated in a C57BL/6 background and then serially back-crossed onto the immunodeficient NOD/SCID strain. Throughout the back-crossing, transgenic mice were identified by Southern blotting. Analysis of more than 500 mice from nine generations of back-crossing revealed that the transgene was passed onto approximately 50% of the animals, with mice of both sexes showing equal distribution of the transgene. RT-PCR revealed that the mRNA was expressed in all tissues tested, including PB, BM, spleen, lung, liver, kidney, heart, and brain (data not shown). Analysis of more than 100 mice from successive generations of back-crossed animals by enzyme-linked immunosorbent assay showed that the level of GM-CSF in serum (100 ng/mL) did not decrease with continued back-crossing to NOD/SCID. Phenotypic analysis of these animals showed no significant differences in the blood counts between transgenic and nontransgenic littermates. We performed colony assays on BM cells from both transgenic and nontransgenic mice, and there was no difference in the frequency or appearance of colonies formed (data not shown). No abnormalities were recorded during routine autopsy of transgenic mice.
Fresh whole PB mononuclear cells from CMML patients were injected into GM-CSF transgenic and littermate nontransgenic NOD/SCID mice exposed to 300 to 350 cGy. Titration of CMML cells showed a clear engraftment in mice transgenic for human GM-CSF (Fig. 4). All mice represented in Figure 4 were analyzed approximately 6 weeks after transplant, and the data are combined from eight individual patients. PB cells from five of these patient samples were injected into both transgenic and nontransgenic mice to compare directly the engraftment in mice producing GM-CSF and normal mice. The level of engraftment was measured by determining the number of cells expressing the myeloid specific antigen CD33. A minimum of 0.5 × 108 cells was required for successful engraftment in NOD/SCID transgenic mice, with no engraftment observed in nontransgenic animals at any of the cell doses tested.
Figure 4Cell dose-dependent engraftment of CMML cells in NOD/SCID mice transgenic for human GM-CSF. Peripheral blood mononuclear cells were injected intravenously into irradiated 8-week-old NOD/SCID mice that were either transgenic for human GM-CSF (filled symbols) or nontransgenic littermates (open symbols). Mice were injected with cells at concentrations between 0.1 and 1.85 × 108 cells per animal. ♥ patient 7, ♦ patient 6, ● 8, ▴ patient 9, ■ all other patients. Isosceles triangle indicates patient 7. Animals were housed for approximately 6 weeks after injections of cells and then analyzed for human cell content. BM cells were stained using antibody specific for human CD33. Each symbol represents the percentage of positive cells as determined by flow cytometric analysis of individual mice. For the mice receiving 1 × 108 cells, statistical analysis showed that there was a statistical improvement on engraftment from the presence of the human GM-CSF transgene (p < 0.05).
To address the question of whether the cells that engrafted in these mice were of leukemic origin or were normal cells from the PB, we tested in parallel the engraftment properties of normal blood cells in human GM-CSF transgenic and nontransgenic NOD/SCID mice. Mononuclear cells were injected into mice at 108 cells per animal and injected mice were housed for 6 weeks. Some of the mice (four transgenic and one nontransgenic) that were given normal human cells died before analysis. Autopsies performed on several of these animals showed that they had empty BM cavities, suggesting that they had not been able to reconstitute their hemopoiesis. In the surviving animals (seven transgenic and three nontransgenic), we were unable to detect any human myeloid cells, although we did see low-level human T cells (as evidenced by CD3 staining) in some animals (data not shown).
We further characterized the engraftment of CMML by performing time course experiments. In particular, we wanted to determine the optimal time to detect engraftment and test the possibility that CMML cells require a longer incubation period to engraft in non–GM-CSF transgenic mice. To test this, we transplanted several mice with 1 × 108 CMML cells and analyzed them 3 to 9 weeks after infusion. Figure 5 shows that prolonging the analysis of transplanted cells beyond 6 weeks does not improve engraftment in nontransgenic mice and that GM-CSF is required for efficient engraftment of CMML cells in vivo.
Figure 5Time course analysis of engraftment levels for CMML peripheral blood cells in transgenic and nontransgenic NOD/SCID mice. Irradiated 8-week-old, NOD/SCID mice were injected with a single infusion of 1 × 108 peripheral blood cells from CMML patients. For this experiment, both transgenic and nontransgenic littermates were used. Mice were housed for 3 to 9 weeks before immunofluorescence analysis for human CD33+ cells. Each symbol represents the percentage of positive cells as determined by flow cytometric analysis of individual mice.
In this study, we present in vitro and in vivo evidence that dependence on GM-CSF is a feature of many CMMLs whose cells show spontaneous colony formation in methylcellulose assays. Furthermore, we show for the first time a mouse model of engraftment of primary CMML cells and describe the optimal conditions required.
Several underlying mechanisms have been implicated in the development and progression of CMML, including mutations of signaling molecules, cytogenetic abnormalities, and various cytokines. In the experiments reported here, we observed spontaneous colony growth in seven of seven CMML BM samples tested. In every CMML case this was reduced by the specific GM-CSF antagonist E21R, indicating that GM-CSF, produced either by the leukemic cells themselves or by accessory cells, was being used for their growth. The levels of inhibition by E21R varied from patient to patient, probably because of dosage differences. Alternatively, other factors may contribute to the residual growth observed, such as hypersensitivity to GM-CSF and responsiveness to other cytokines.
Data from our study points to an important role of GM-CSF in the pathogenesis of certain CMML and may be related to the fact that a significant proportion of CMML patients have elevated levels of GM-CSF in the serum [
]. A parallel can be drawn to JMML with respect to the effects of E21R, which showed that antagonizing GM-CSF in vitro abrogated leukemic cell growth [
Inhibition of proliferation and induction of apoptosis in juvenile myelomonocytic leukemic cells by the granulocyte-macrophage colony-stimulating factor analogue E21R.
]. Thus, two different leukemias within the new category of MDS/MPD require GM-CSF for growth in vitro and in vivo and show an important biologic difference with MPD induced by tyrosine kinase fusion oncogenes. The latter has been shown in an experimental murine model not to require GM-CSF for their development [
Induction of myeloproliferative disease in mice by tyrosine kinase fusion oncogenes does not require granulocyte-macrophage colony-stimulating factor or interleukin-3.
During the past 10 years, several mouse models have been developed for their potential to engraft human leukemic cells. Original studies investigating the engraftment of human cells in SCID mice (without the additional NOD mutation) showed that it was necessary to have human cytokines for the engraftment to succeed. Daily injections of human cytokines later were improved by either the generation of transgenic mice [
]. A comparison of several strains of mice by Dick et al. concluded that the NOD/SCID strain, which lacks T and B cells and has reduced natural killer cell, complement, and macrophage activity [
High level engraftment of NOD/SCID mice by primitive normal and leukaemic hematopoietic cells from patients with chronic myeloid leukemia in chronic phase.
]. In general, NOD/SCID mice do not require the addition of exogenous cytokines for engraftment of these leukemic samples. For CMML, however, no mouse experimental model has been developed to date.
Initially we were unable to engraft CMML cells in NOD/SCID mice. For this reason we generated NOD/SCID mice that are transgenic for human GM-CSF. Human GM-CSF transgenic mice were back-crossed for seven to nine generations with NOD/SCID mice so that the immunodeficient phenotype could be maintained. We found that it is necessary to inject 5 to 10 × 107 CMML cells (which is equivalent to 0.2–3.6 × 106 CD34+ cells) to obtain detectable engraftment in the transgenic NOD/SCID mice (Fig. 4). This is comparable to work from Bonnet and Dick [
], who showed that the frequency of the leukemia-initiating cell (LIC) was 1 per 104 CD34+ cells or between 0.2 and 200 per 5 × 106 mononuclear cells. In addition, GM-CSF may have further facilitated engraftment through its role in promoting cell survival. To verify that it was leukemic cells being detected in the mouse BM, we attempted to engraft normal human PB in the GM-CSF transgenic NOD/SCID mice. Under our conditions of transplantation we were unable to detect any human myeloid cells in mice injected with normal blood, strongly suggesting that engraftment of CMML samples is due to the leukemic cells. To formally prove this it would be necessary to use CMML samples with genetic abnormalities. Optimal levels of CMML engraftment were found 6 weeks after infusion, with no enhancement after this time. Injecting cells from the same patient into normal vs transgenic mice showed an advantage for the cells injected into mice that produced human GM-CSF. Five patients were tested in this way, and the data are shown in Figure 4, with individual patients depicted using different symbols.
This murine model of CMML may be useful for studying its underlying mechanisms and may form the basis of a future model in which therapeutic drugs for CMML can be tested. With respect to GM-CSF, this cytokine may be blocked by neutralizing antibodies or by the specific antagonist E21R [
]. Additionally, GM-CSF fused to toxins may be valuable in attacking cells expressing the GM-CSF receptor; however, the effects may not be restricted to leukemic cells. Diptheria toxin fused to GM-CSF (DT388–GM-CSF) has been shown to kill CMML cells [
Diptheria toxin fused to granulocyte-macrophage colony-stimulating factor is toxic to blasts from patients with juvenile myelomonocytic leukemia and chronic myelomonocytic leukemia.
], but some general toxicity could be seen. In the development of another potential therapy, GM-CSF was fused to Pseudomonas exotoxin (GM-CSF–PE40). This was shown to be functionally active, but it was not cytotoxic to leukemic cell lines, probably because the PE moiety is unable to be internalized as required for function [
The experiments described here indicate that GM-CSF can significantly contribute to the pathogenesis of certain cases of CMML; however, they do not distinguish whether GM-CSF is necessary for the initiation or progression of the disease. In addition, certain MPDs induced by tyrosine kinase fusion proteins may not have an absolute requirement for GM-CSF [
]. Nevertheless, the blocking of GM-CSF action in vivo may provide an effective strategy to control and ameliorate the progression of this leukemia in some CMML patients.
Acknowledgements
Supported by a University of Adelaide Research Associateship (H.R.), the Anti-Cancer Foundation of South Australia (H.R. and P.B.), and BresaGen Ltd. (M.L.).
Transformation of hematopoietic cell lines to growth-factor independence and induction of a fatal myelo- and lymphoproliferative disease in mice by retrovirally transduced TEL/JAK2 fusion genes.
The TEL/platelet-derived growth factor b receptor (PDGFβR) fusion in chronic myelomonocytic leukemia is a transforming protein that self-associates and activates PDGFbR kinase-dependent signaling pathways.
Interleukin-10 inhibits spontaneous colony-forming unit growth from human peripheral blood mononuclear cells by suppression of endogenous granulocyte-macrophage colony-stimulating factor release.
Inhibition of granulocyte-macrophage colony-stimulating factor prevents dissemination and induces remission of juvenile myelomonocytic leukemia in engrafted immunodeficient mice.
Barry SC, Bagley CJ, Phillips J, et al. (1994) Two contiguous residues in human interleukin-3, Asp21 and Glu22, selectively interact with the α- and β-chains of its receptor and participate in function. J
Hogan B, Beddington R, Constantini F, Lacy E (1994) Manipulating the Mouse Embryo. A Laboratory Manual. New York: Cold Spring Harbor Laboratory Press, p. 217
Monoclonal antibody 7G3 recognizes the N-terminal domain of the human interleukin-3 (IL-3) receptor α-chain and functions as a specific IL-3 receptor antagonist.
Inhibition of proliferation and induction of apoptosis in juvenile myelomonocytic leukemic cells by the granulocyte-macrophage colony-stimulating factor analogue E21R.
Induction of myeloproliferative disease in mice by tyrosine kinase fusion oncogenes does not require granulocyte-macrophage colony-stimulating factor or interleukin-3.
High level engraftment of NOD/SCID mice by primitive normal and leukaemic hematopoietic cells from patients with chronic myeloid leukemia in chronic phase.
Diptheria toxin fused to granulocyte-macrophage colony-stimulating factor is toxic to blasts from patients with juvenile myelomonocytic leukemia and chronic myelomonocytic leukemia.
00903-7&id=gr1.jpg){kind=link}
00903-7&id=gr2.jpg){kind=link}
00903-7&id=gr3.jpg){kind=link}
00903-7&id=gr4.gif){kind=link}
00903-7&id=gr5.gif){kind=link}