Growth autonomy and lineage switching in BCR-ABL ... - Nature

Leukemia (2004) 18, 1006–1012 & 2004 Nature Publishing Group All rights reserved 0887-6924/04 $25.00 www.nature.com/leu

Growth autonomy and lineage switching in BCR-ABL-transduced human cord blood cells depend on different functional domains of BCR-ABL Y Chalandon1,2,6, X Jiang1,6, S Loutet1, AC Eaves1,3,4 and CJ Eaves1,3,4,5 1 Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver, Canada; 2Hematology Division, University Hospital of Geneva, Geneva, Switzerland; 3Department of Medicine, University of British Columbia, Vancouver, BC, Canada; 4Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada; and 5Department of Medical Genetics, University of British Columbia, Vancouver, BC, Canada

The tyrosine kinase activity of p210BCR-ABL is essential to its leukemogenic potential, but the role of other functional domains in primary human hematopoietic cells has not been previously investigated. Here we show that infection of normal human CD34 þ cord blood (CB) cells with a retroviral vector encoding p210BCR-ABL rapidly activates a factor-independent phenotype and autocrine interleukin-3/granulocyte colonystimulating factor/erythropoietin production in the transduced cells. These changes are characteristic of primitive chronic myeloid leukemic (CML) cells and are important to the leukemogenicity of BCR-ABL-transduced murine hematopoietic stem cells. When BCR-ABL-transduced human CB cells were incubated with imatinib mesylate, an inhibitor of the p210BCR-ABL kinase, or when human CB cells were transduced with a BCR-ABL cDNA lacking the SH2 domain (p210DSH2), factor independence was significantly reduced. In contrast, deletion of the SH2 domain had little impact on the p210BCR-ABL kinase-dependent promotion of erythropoietic differentiation also seen immediately following the BCR-ABL transduction of primitive human CB cells, but not in naturally occurring CML. Thus, p210BCR-ABL has distinct biological effects in primary human hematopoietic cells, which variably mimic features of human CML, and activation of these changes can show different dependencies on the integrity of the SH1 and SH2 domains of p210BCR-ABL. Leukemia (2004) 18, 1006–1012. doi:10.1038/sj.leu.2403335 Published online 11 March 2004 Keywords: BCR-ABL; chronic myeloid leukemia; autocrine; SH2 domain; imatinib mesylate

Introduction The BCR-ABL oncogenes comprise a family of fusion genes key to the genesis of a group of related myeloproliferative disorders.1 These disorders include chronic myeloid leukemia (CML) and a proportion of cases that present as acute leukemia without evidence of a preceding chronic phase involving the myeloid lineages. The BCR-ABL oncogene associated with classical CML produces a 210 kDa oncoprotein (p210BCR-ABL)2 that has a constitutively active tyrosine kinase encoded by the Src homology 1 (SH1) domain3 and an abnormal cytoplasmic location.4 Both of these features are key to the transforming activity of p210BCR-ABL in model systems.5–10 The importance of the kinase function of p210BCR-ABL in human BCR-ABL þ disease Correspondence: Dr CJ Eaves, Terry Fox Laboratory, 601 West 10th Avenue, Vancouver, BC, Canada V5Z 1L3; Fax: þ 604 877 0712; E-mail: [email protected] Supported by grants from the National Cancer Institute of Canada with funds from the Canadian Cancer Society and the Terry Fox Run and a grant from Genome BC. Y Chalandon was supported by a Terwindt Foundation Fellowship and S Loutet held a British Columbia Cancer Foundation Summer Studentship 6 Both these authors contributed equally to this work Received 20 January 2004; accepted 5 February 2004; Published online 11 March 2004

is indicated by the potent antileukemic effect of imatinib mesylate (Gleevec, STI571),11–13 an agent that selectively blocks the kinase activity of p210BCR-ABL.9 Studies of the in vivo leukemogenic activity of primary mouse bone marrow (BM) cells transduced with P210BCR-ABL constructs lacking the SH2 have yielded contradictory findings as to the importance of this domain.14–16 In primary human hematopoietic cells, very little is known about the importance of different domains of p210BCR-ABL with regard to its transforming activity. Primitive (CD34 þ ) cells from CML patients are characterized by factor-independent growth and differentiation in vitro.17–20 At the level of primitive CML progenitors, this abnormal behavior is due, at least in part, to the acquisition by the neoplastic cells of an autocrine interleukin-3 (IL-3)/granulocyte colony-stimulating factor (G-CSF) growth mechanism.21 P210BCR-ABL-induced production of IL-3 by hematopoietic cells was first demonstrated in various immortalized, factor-dependent cell lines,22–24 and later confirmed in BCR-ABL-transduced primary mouse BM cells25–27 and embryonic stem cells.28 A subsequent study showed that the lethal leukemias obtained following the transplantation of large numbers of BCR-ABLtransduced IL3/ or granulocyte–macrophage colony-stimulating factor (GM-CSF)/ mouse BM cells raised questions about the importance of autocrine factor production to the leukemogenic potential of primitive CML cells.29 However, more recent experiments with the same model showed that the generation of leukemia was critically dependent on the persistence of an autocrine IL-3 mechanism when the production of other growth factors was attenuated by short-term passage of the BCR-ABLtransduced cells in culture prior to the transplantation of reduced numbers of leukemic stem cells.27 We have also obtained evidence of an enhanced production of erythropoietin (Epo) transcripts in primary human CML progenitors (Jiang X, Fujisaki T, Berger M, Eaves A, Eaves C. Blood 1998; 92 (Suppl. 1): 254 (abstract)), which could contribute to the observed ability of some terminally differentiating CML erythroid precursors to mature in vitro in the absence of added Epo.17 Forced expression of BCR-ABL in primary human hematopoietic cells offers a powerful strategy for creating new models of human CML to interrogate mechanisms of disease genesis and the effects of novel therapeutics. To date, the use of this approach has been limited, but perturbations in the adhesive properties of the transduced cells30 and their lineage commitment31 have been documented in transduced human cord blood (CB) cells. Here we show that these cells, like their murine counterparts,27 rapidly acquire a growth factor-independent phenotype concomitant with the activation of IL-3, G-CSF and Epo expression. In addition, we show that inhibition of the SH1 (kinase) domain and deletion of the SH2 domain of p210 BCR-ABL have different effects on the BCR-ABL-induced perturbations of growth and differentiation exhibited by primitive human hematopoietic cells immediately post-transduction.

BCR-ABL domain-specific effects on human cells Y Chalandon et al

1007 Materials and methods

Virus production The MSCV-BCR-ABL-IRES-GFP vector (P210) encoding a fulllength b3a2 p210BCR-ABL cDNA and the control MSCV-IRESGFP (MIG) virus used here have both been described previously.27 The MSCV-BCR-ABLDSH2-IRES-GFP vector (P210DSH2) containing a mutant (SH2-deleted) BCR-ABL cDNA (from R Van Etten, Center for Blood Research, Boston, MA, USA) was constructed by replacing the original BCR-ABL cDNA with the mutant cDNA in the same vector and was further verified by restriction enzyme digestion analysis (Figure 1). Virus preparations were obtained by transfecting amphotropic Phoenix packaging cells32 cultured in DMEM plus 10% fetal calf serum (FCS, StemCell Technologies, Vancouver, BC, Canada) using CaPO4. Virus-containing medium (VCM) was harvested 32–72 h later,33 filtered through a 0.45 mm filter and used immediately for transduction. Titers of the P210, P210DSH2 and control (MIG) virus were 2–5, 2–4 and 5–9 105 infectious units/ml, respectively, as assessed by green fluorescence protein (GFP) expression in NIH 3T3 cells 2 days after their exposure to limiting dilutions of VCM.

Cells CB cells were obtained from consenting mothers undergoing cesarean delivery of normal, full-term infants. Low-density (o1.077 g/ml) cells were isolated using Ficoll–Hypaque (Amersham Pharmacia Biotech, Baie d’Urfe, Quebec, Canada) and a population enriched in CD34 þ cells (7273% CD34 þ , n ¼ 12) obtained by immunomagnetic removal of lineage marker þ (lin þ ) cells on columns (StemSept, StemCell Technologies), as recommended by the supplier.

Retroviral transduction Cells to be transduced were cultured at 1–2 105/ml for 48 h in Iscove’s minimal essential Dulbecco’s medium (IMDM, StemCell Technologies) supplemented with complete serum-free medium (SFM) containing the following five purified recombinant human growth factors: Flt-3 ligand (FL, 100 ng/ml, Immunex Corporation, Seattle, WA, USA), steel factor (SF,

Figure 1 Retroviral vectors. Schematic diagram of the MSCVbased retroviral vectors used. These result in expression of GFP only, GFP plus the full-length BCR-ABL cDNA, or GFP plus a BCR-ABL cDNA with a deleted SH2 domain. LTR ¼ long terminal repeat (derived from MSCV), IRES ¼ internal ribosomal entry site.

100 ng/ml, produced in the Terry Fox Laboratory), IL-3 (20 ng/ ml, Novartis, Basel, Switzerland), IL-6 (20 ng/ml, Cangene, Mississauga, ON, Canada) and G-CSF (20 ng/ml, StemCell Technologies). The SFM consisted of IMDM supplemented with a combination of bovine serum albumin, insulin and transferrin (BITt, StemCell Technologies), 40 mg/ml low-density lipoproteins and 104 mmol/l 2-mercaptoethanol (both from Sigma Chemicals, St Louis, MO, USA). The cells were then resuspended in filtered VCM supplemented with the same five growth factors and 5 mg/ml protamine sulfate (Sigma). The cells were then incubated at 371C for 12 h in fibronectin-coated Petri dishes previously ‘loaded’ twice with VCM (each time for 30 min). This infection procedure was repeated twice (three 12h exposures to fresh virus) and the cells finally transferred to fresh SFM plus the same five growth factors but no VCM and incubated for a further 48 h prior to harvesting and further analysis. In experiments in which effects of imatinib mesylate (Gleevec/STI571, Novartis) were investigated, 0.25 mmol/l was added to the medium from the time of the first exposure to virus.

Flow cytometry Viable GFP þ and GFP CD34 þ CB cells were isolated at a purity of 499% using a three-laser FACStar þ cell sorter (Becton Dickinson, San Jose, CA, USA) and collected in eppendorf tubes containing complete SFM as described previously.34

Cultures Transduced cells were cultured in IMDM with complete SFM with (or without) the same five growth factor cocktail used in the transduction protocol. Colony-forming cell (CFC) assays were performed in SFM-containing methylcellulose cultures (Methocult H4100, StemCell Technologies) supplemented or not with human SF (50 ng/ml) and 20 ng/ml each of human IL-3, IL-6, GCSF and GM-CSF73 U/ml human Epo (StemCell Technologies) as described previously.31

Reverse transcriptase (RT)–PCR and Southern blot analyses Total RNA was extracted from aliquots of 104 cells as described previously.35 The RT reaction was performed in 20 ml volumes with superscript II RNase H RT (GIBCO/BRL, Rockville, MD, USA) using random hexamer oligonucleotides as primers (Amersham Pharmacia). A measure of 5 ml of the RT reaction was then used for PCR amplification in 50 ml volumes of 1 PCR buffer (GIBCO/BRL) containing 20 mmol/l Tris-HCl (pH 8.4), 50 mmol/l KCl, 2 mmol/l MgCl2, 200 mmol/l of each dNTP (Amersham Pharmacia), 1 U of Taq polymerase and 10 pmol/l of specific primers for BCR-ABL (50 -CAGGGTGCACAGCCGCAACGGCAA-30 and 50 -GTCCAGCGAGAAGGTTTTCCTTGGA-30 ), GAPDH (50 -GTCTTCACCACCATGGAGAAGG-30 and 50 -GCCTGCTTCACCACCTTCTTGA-30 ), IL-3 (50 -GCTCCCATGACCCAGACAACGTCC-30 and 50 -CAGATAGAACGTCAGTT TCCTCCG-30 ), G-CSF (50 -CTCTGGACAGTGCAGGAAGCCA CC-30 and 50 -GCTGGGCAAGGTGGCGTAGAACGC-30 ), thrombopoietin (TPO, 50 -GGCCAGAATGGAGCTGACTGAATTG-30 and 50 -TCCTACAAGCATCAGGAAACGCACC-30 ), SF (5’-GGAT CTGCAGGAATCGTGTGACTA-3’ and 5’-CTTCAGGAGTAAAGAGCCTGGGTT-3’), Epo (5’-ATCACGACGGGCTGTGCTGAA CAC-3’ and 5’-GGGAGATGGCTTCCTTCTGGGCTC-3’), IL-6 Leukemia


1008 (5’-CCACACAGACAGCCACTCACCTCT-3’ and 5’-CTGCAGGAACTGGATCAGGACTTT-3’) and GM-CSF (5’-CAGGAGGC CCGGCGTCTCCTGAAC-3’ and 5’-ACAAGCAGAAAGTCCTTCAGGTTC-3’). These generated DNA fragments of 374 bp (BCRABL), 493 bp (GAPDH), 345 bp (IL-3), 547 bp (G-CSF), 495 bp (TPO), 335 bp (SF), 289 bp (Epo), 341 bp (IL-6) and 300 bp (GMCSF). A total of 35 cycles of amplification at 941C for 30 s, 621C for 1 min and 721C for 1 min were then performed. A measure of 10 ml aliquots of the PCR product were then electrophoresed, transferred to a nylon membrane and probed with cDNA probes for human IL-3, G-CSF, GM-CSF, IL-6, SF, Epo and GAPDH (from R Kay and K Humphries, Terry Fox Laboratory, Vancouver, BC, Canada), and for BCR-ABL (from J Griffin, Dana-Farber Cancer Center, Boston, MA, USA) as described previously.21 For the detection of human TPO cDNA, an internal sequence (50 -CCGAGTCCTCAGTAAACTGCTTCG-30 ) was used as a hybridization probe.

Western analyses A total of 105 cells per treatment group were lysed in phosphorylation solubilization buffer (PSB; 50 mmol/l Hepes buffer, 100 mmol/l NaF, 10 mmol/l sodium pyrophosphate, 2 mmol/l Na3VO4, 4 mmol/l EDTA) containing 0.5% Nonidet P-40, and Western analyses were performed as described previously using a mouse monoclonal anti-Abl antibody corresponding to the carboxyl region of Abl (Ab-3, Oncogene Research Products, Boston, MA, USA) and an antiphosphotyrosine antibody (4G10, Upstate Biotechnology, Lake Placid, NY, USA).20

Rapid acquisition of growth autonomy by BCR-ABL-transduced human CB cells

Figure 2 Growth kinetics of BCR-ABL-expressing human CD34 þ CB cells. (a) P210-transduced CD34 þ CB cells display immediate SH2 domain-dependent growth factor independence. Serum-free cultures were initiated with 1–5 104 FACS-sorted P210 (triangles)-, P210DSH2 (circles)- or MIG (squares)-transduced (GFP þ ) CD34 þ CB cells and then incubated in the presence (solid symbols) or absence (open symbols) of added growth factors (GF ¼ FL, SF, IL-3, IL-6 and GCSF). Total viable (trypan blue-excluding) cells were measured after 7 and 14 days. Results are from three independent experiments. (b) Imatinib mesylate inhibits the growth of P210-transduced CD34 þ CB cells. Same experiments as in (a) but showing viable cell counts from cultures of P210 (triangles)- or MIG (squares)-transduced (GFP þ ) CD34 þ CB cells cultured in the presence (open symbols) or absence (solid symbols) of 0.25 mmol/l imatinib mesylate. All cultures contained FL, SF, IL-3, IL-6 and G-CSF except one group of P210transduced (GFP þ ) CD34 þ CB cells (open diamonds) that were cultured without growth factors in the presence of imatinib mesylate.

In a first set of experiments, lin CB cells (7273% CD34 þ ) were transduced with either the P210 or the control (MIG) vector (Figure 1) after 2 days of growth factor prestimulation (with FL plus SF plus IL-3 plus IL-6 plus G-CSF) and then assessed for their ability to proliferate and differentiate in SFM in the presence or absence of added growth factors, both in suspension culture and in semisolid medium. The initial proportion of cells that were GFP þ (as assessed after 2 days in suspension cultures containing the same potent five growth factors used in the transduction protocol) was 1171% for P210 and 2874% for MIG (n ¼ 6). This result is consistent with the 2to 3-fold higher titer of the MIG virus preparations as assessed on 3T3 cells. Interestingly, within the CD34 þ subset of CB cells, the % GFP þ cells 2 days post-transduction was the same for both vectors (772 and 871%), perhaps due to a selectively enhanced early rate of amplification of an initially lower fraction of P210-expressing CD34 þ cells.31 The rapidity of the effect of p210BCR-ABL on the growth rate of the transduced CB cells was seen also in the present experiments (Figure 2a), where a 50-fold greater net expansion of the P210-transduced (sorted GFP þ )

cells vs MIG-transduced (sorted GFP þ ) cells over a 14-day period can be seen (100 000-fold vs 2000-fold, Po0.05, n ¼ 3). Also shown in Figure 2a is a comparison of the changes in cell numbers seen in the parallel cultures initiated with the same sorted GFP þ cells to which no growth factors were added. Under these conditions, the P210-transduced CD34 þ cells expanded at a rate almost equivalent to that seen when the five growth factor cocktail was added (B2-fold less). In contrast, in the absence of growth factors, the control cells died rapidly. Similarly, when aliquots of the same transduced cells were plated immediately post-transduction in methylcellulose, many erythroid and granulocyte–macrophage or mixed colonies were obtained from the FACS-sorted P210-transduced (GFP þ ) cells in the absence of added growth factors (overall, at 17% the efficiency as with growth factors, Table 1). In contrast, the control cells formed colonies only in cultures to which growth factors were added.

Statistical analysis Results are shown as the mean7s.e.m. of values obtained in at least three independent experiments. Differences between groups were assessed using the two-tailed Student’s t-test.

Results

Leukemia


1009 Table 1 CB cells transduced with a P210DSH2 virus show a marked loss of growth factor independence with minimal effects on the induction of erythropoiesis caused by P210 p210+GF CFU-E BFU-E CFU-GEMM CFU-GM Total CFC

a

774 3807110a 60715a 40710a 4907130

p210-GF b

4078 4075b 472 171 85710b

DSH2+GF a

2575 240750a 90730a 60710a 410770

DSH2-GF c

372 473c 070 070 774c

MIG+GF

MIG-GF

070 95715 170720 180740 450740

070 070 070 070 070

CFC assays were performed 2 days after the completion of the transduction protocol to allow full expression of the transduced genes prior to plating in colony assays. Mean7s.e.m. values from three independent experiments are expressed per 100 pretransduction cells to allow the CFC expansion due to growth factor stimulation for 6 days to be normalized. GF ¼ growth factors. a Significantly different from the MIG-control-transduced cells assayed in the presence of growth factors (Po0.05). b Significantly different from the MIG-transduced cells assayed in the absence of growth factors (Po0.05). c Significantly different from the P210-transduced cells assayed in the absence of growth factors (Po0.05).

Comparison of the different types of colonies generated from the P210- and control-transduced cells showed a significant decrease in the number of granulopoietic colonies and a correspondingly increased number of erythroid colonies produced by the P210-transduced cells. We have previously shown that this lineage switching response is due to an ability of forced expression of P210 in uncommitted primary human progenitors to favor selection of the erythroid lineage, and a re-activation of erythroid potential and suppression of granulopoietic potential in progenitors otherwise thought to be already restricted to the granulocyte–macrophage pathway.31

Rapid induction of expression of selected growth factors in P210-transduced human CD34 þ CB cells To investigate the basis of the growth factor autonomy displayed by P210-transduced CD34 þ CB cells, we looked for evidence of activation of the same autocrine mechanisms operative in primary CD34 þ CML progenitors. These include the production of IL-3, G-CSF and Epo but not SF, TPO, GM-CSF or IL-621 (Jiang X, Fujisaki T, Berger M, Eaves A, Eaves C. Blood 1998; 92 (Suppl. 1): 254 (abstract)). Examination by RT–PCR of RNA extracts from FACS-sorted GFP þ cells transduced with either P210 or MIG and then cultured in suspension for 2, 3 and 7 days showed a rapid and selective induction of IL-3, G-CSF and Epo transcripts in the P210-transduced cells. Results representative of three similar experiments are shown in Figure 3. In these experiments, both the P210- and MIG-transduced cells were cultured with growth factors to minimize selection of P210transduced cells with the greatest factor independence and to minimize apoptosis in the control cultures. Transcripts for IL-3 and G-CSF were also consistently detectable in the MIGtransduced cells by day 7 of culture, presumably due to the generation by that time of mature cell types known to produce these growth factors (eg, mast cells and macrophages). When the same RNA extracts were analyzed for the presence of SF and IL-6 transcripts, only occasional and sporadic evidence of these was seen in both control- and P210-transduced cells (data not shown). A similar search for GM-CSF or TPO transcripts yielded totally negative results in all samples (data not shown).

Interference with the kinase activity of p210BCR-ABL abrogates the growth factor independence of transduced human CB cells We next asked whether the p210BCR-ABL kinase activity was critical to the growth autonomy and autocrine factor production

Figure 3 Activated growth factor gene expression in P210transduced (GFP þ ) CB cells and its inhibition by imatinib mesylate. Southern blots showing RT–PCR detection of IL-3, G-CSF and Epo transcripts in RNA extracted from MIG- or P210-transduced CD34 þ CB cells cultured for 2, 3 or 7 days immediately post-transduction in SFM plus growth factors (FL, SF, IL-3, IL-6 and G-CSF) with or without 0.25 mmol/l imatinib mesylate, as indicated. Viable GFP þ cells were isolated by FACS prior to RNA extraction. GAPDH transcripts are present in all samples. Lane 1 contained no RT (negative control for RT). Lane 2 contained water (negative control for PCR). Lanes 3–14 contained amplified products from the MIG- and P210-transduced CB cells from a representative experiment. Lane 15 contained amplified products from K562 cells (positive control for BCR-ABL). Lane 16 contained amplified products from either a mouse fibroblast cell line (M210B4 cells) engineered to express human IL-3 and G-CSF,51 or CHOK-1 cells (CHO cells engineered to express human Epo). The expected sizes of each PCR product are indicated.

characteristic of P210-transduced CD34 þ CB cells. To address these questions, we cultured additional aliquots of cells of P210transduced cells with or without growth factors in the presence of a low ‘nontoxic’ concentration (0.25 mmol/l) of imatinib mesylate that we had previously found to be sufficient to normalize the growth rate and differentiation of P210-transduced human CB cells cultured in the presence of growth factors.31 This effect was confirmed in the present experiments (Figure 2b). In contrast, in the absence of added growth factors, 0.25 mmol/l of imatinib mesylate caused a pronounced reduction in the growth of P210-transduced cells (104 times fewer viable cells present after 14 days, as indicated by a comparison Leukemia


1010 of the data in Figure 2b with the data in Figure 2a for P210transduced cells cultured without growth factors in the absence of the inhibitor). However, under these conditions, a complete restoration of the degree of factor dependence exhibited by control (MIG-transduced) cells was not achieved. To ascertain whether the marked effect of imatinib mesylate on the loss of factor independence by P210-transduced CB cells might be associated with a P210-kinase-dependent ability to activate autocrine growth factor production, transduced cells that had been cultured with 0.25 mmol/l imatinib mesylate (plus exogenously provided growth factors) were examined for the presence of growth factor transcripts. As shown in the right-hand lanes in Figure 3, exposure to imatinib mesylate eliminated the early induction of IL-3 and Epo transcripts seen in control P210transduced cells (cells cultured without the inhibitor shown on the left-hand side of Figure 3), whereas the activation of G-CSF production, at least with this dose of inhibitor, appeared unaffected.

Deletion of the SH2 domain differentially affects the growth factor independence and altered commitment behavior of P210-transduced human CB cells We then asked whether the SH2 domain of p210BCR-ABL was also important to the acquisition by primitive human hematopoietic cells of a factor-independent phenotype. For this, we produced a retroviral vector capable of infecting human cells that was identical to the P210 vector but with the SH2 domain deleted (P210DSH2, Figure 1). The titer of the P210DSH2 virus as assessed on 3T3 cells was similar to that of the intact P210 virus, and lin CB cells were also transduced with the P210DSH2 virus at a similar efficiency (1171% GFP þ total cells; 771% GFP þ CD34 þ cells, n ¼ 6) as described above for P210. However, the growth of P210DSH2-transduced CB cells in 14-day suspension cultures with growth factors was, unlike that exhibited by cells transduced with P210, not accelerated relative to the MIG-transduced controls (Figure 2a). Furthermore, in the absence of added growth factors, the P210DSH2transduced CB cells showed significantly diminished, albeit not completely abrogated, growth autonomy (100-fold less net expansion than the P210-transduced cells cultured without growth factors and 16-fold less net expansion than the P210DSH2-transduced cells cultured with growth factors, but still 41000-fold greater net expansion than the MIG-transduced cells cultured without growth factors, Po0.05). A marked but incomplete restoration of factor dependence was also seen in the assays of the P210DSH2-transduced CB cells in semisolid medium (1.7% factor-independent colony formation vs 17% by P210, Table 1). However, this was not accompanied by more than a marginal change in the altered differentiation potential exhibited by the P210-transduced cells. Thus the significantly increased numbers of erythroid colonies and decreased numbers of granulocyte–macrophage colonies obtained from P210-transduced CB cells was also a feature of the assays of the P210DSH2-transduced CB cells, although there was evidence of a slightly diminished switch in the P210DSH2transduced cells. This finding contrasts with the strongly normalizing effect on lineage commitment that we previously reported is obtained when P210-transduced human CB cells are assayed in the presence of 0.25 mmol imatinib mesylate.31 Because previous studies have shown that deletion of the SH2 domain can reduce the kinase activity of P210BCR-ABL in cell lines,24,36,37 it was of interest to compare the levels of tyrosinephosphorylated proteins in primary P210- and P210DSH2Leukemia

transduced human CB cells. Accordingly, FACS-sorted GFP þ CD34 þ cells were cultured for 7 days in suspension culture with growth factors and then lysates prepared for Western analysis. As illustrated in Figure 4, the level of expression of the mutant p210BCR-ABLDSH2 in CB cells transduced with P210DSH2, as indicated by blotting with an anti-c-abl antibody, was consistently lower than the level of expression of intact p210BCR-ABL in P210-transduced aliquots of the same target cells. However, both were still consistently much higher than the levels of P210 seen in K562 cells – an established human leukemic cell line with multiple Ph chromosomes.38 This presumably reflects a greater potency of the long terminal repeat promoter of the vector in transduced cells by comparison to the BCR promoter in CML cells. Prior blotting of the same cell lysates with an antiphosphotyrosine antibody showed the levels of most of the tyrosine-phosphorylated proteins in the P210DSH2-expressing CB cells to be reduced relative to the P210-transduced CB cells, but higher than in K562 cells. However, some heterogeneity in the effect of the SH2 deletion on the reduction in phosphorylated proteins of different molecular weights was seen, as has been previously reported for other cell targets.24,36,37

Discussion In this study, we have used a gene transfer strategy to demonstrate that the IL-3/G-CSF/Epo autocrine mechanism of growth autonomy characteristic of CD34 þ human CML cells21 (Jiang X, Fujisaki T, Berger M, Eaves A, Eaves C. Blood 1998; 92 (Suppl. 1): 254 (abstract)) is an immediate and, in the case of IL3 and Epo, a p210BCR-ABL kinase-dependent effect of the expression of BCR-ABL in primitive human hematopoietic cells. These findings provide strong support for the view that the autocrine mechanism seen in primitive CML cells is a direct

Figure 4 Decreased expression and phosphorylation of P210DSH2-transduced CB cells as compared to wild-type P210transduced CB cells. MIG-, P210- and P210DSH2-transduced CB cells were cultured for 7 days in SFM in the presence of FL, SF, IL-3, IL-6 and G-CSF and then Western blot analysis was performed on the total cell lysates. Results are from a representative of three experiments. The expression and phosphorylation of p210BCR-ABL in the cultured P210and P210DSH2-transduced CB cells was greatly increased as compared to K562 cells as shown by first probing with an antiphosphotyrosine antibody (lower panel) and then an anti-ABL antibody (upper panel). The position of p210BCR-ABL and the endogenous p145c-ABL (c-ABL, endogenous loading control) is indicated. Total cell lysates from MIG-transduced CB cells and from the K562 cell line were used as negative and positive controls, respectively.


consequence of their expression of p210BCR-ABL. We also observed that transduction of primitive human CB cells with an SH2 domain-deficient BCR-ABL cDNA significantly compromised their ability to grow autonomously. Previous studies of the role of the SH2 domain in the induction of leukemogenic activity, growth factor autonomy and the activation of autocrine mechanisms by p210BCR-ABL in either factor-dependent murine cell lines or primary murine hematopoietic cells have yielded contradictory results. In some cases, this domain was dispensible,15,16,24,36,37,39 but not in others.14,24,40 One interpretation of these findings is that p210BCR-ABL may confer growth factor autonomy by multiple mechanisms, of which autocrine growth factor production may be only one. Although this possibility was not specifically addressed in the present study, the incomplete restoration of normal growth factor dependence in the P210DSH2-transduced cells is consistent with such a hypothesis. Regardless of the final explanation, our present findings with the P210DSH2 mutant re-emphasize the importance of the cellular context in which the consequences of BCR-ABL expression are investigated, as confirmed for other downstream effects of p210BCR-ABL.26,27,41 As additionally demonstrated here, this includes the failure of P210BCR-ABL to activate production of GM-CSF in primitive human hematopoietic cells in contrast to their murine counterparts.25,27 We also documented an enhanced production of Epo transcripts in naı¨ve human progenitors forced to express p210BCR-ABL and showed that this response was dependent on the kinase activity of this oncoprotein. The neoplastic erythroid progenitors from many CML patients display a sufficiently reduced requirement for Epo to allow the formation of recognizable colonies of hemoglobinized erythroblasts in vitro under conditions of Epo deprivation that do not support the growth and maturation of their normal counterparts,17 in spite of the ability of the latter to produce low levels of Epo.42,43 The present experiments suggest the BCR-ABL-mediated activation of another autocrine loop in human hematopoietic targets with specificity for terminally differentiating erythroid cells. However, alternative mechanisms that do not involve Epo receptor activation may also be operative.44 The rapidity of the activation of expression of IL-3, G-CSF and Epo in BCR-ABL-transduced human CD34 þ CB cells raises questions as to whether such an autocrine loop may play a role in the enhanced commitment to erythropoiesis that these cells exhibit immediately post-transduction. IL-345 and G-CSF46 are well known for their ability to promote the expansion of primitive erythropoietic cells, as is the role of Epo as a critical regulator of the terminal stages of this pathway.47 However, convincing evidences of cytokine- (or cytokine receptor-) directed changes in the erythroid commitment of pluripotent progenitors, or of cells already restricted to the granulocyte– macrophage pathway, have not been reported. Indeed, most experimental strategies used to date to test such possibilities have yielded negative results.48–50 Here we show that the autonomous growth potential acquired by BCR-ABL-transduced primary human hematopoietic cells, like their enhanced generation of erythroid progeny at the expense of granulopoietic cells,31 is highly dependent on the kinase activity of the p210BCR-ABL oncoprotein. In contrast, these two responses appear to be differentially affected by the removal of the SH2 domain from the P210BCR-ABL cDNA used to transduce the cells. The SH2 deletion mutant significantly reduced their factor independence but had little impact on their perturbed commitment behavior. Taken together, these results indicate that qualitatively (or at least quantitatively) different mechanisms are required to mediate these changes and that the events that

1011 perturb the process of lineage restriction are unlikely to be downstream of a p210BCR-ABL-activated autocrine mechanism. In summary, it appears that forced expression of high levels of p210BCR-ABL in primitive normal human hematopoietic target cells can initiate a spectrum of biological changes that are differentially dependent on the integrity of different domains of this oncoprotein. The fact that some of these changes (eg, activation of the autocrine production of IL-3, G-CSF and Epo) are also characteristic of CML progenitors invites further exploitation of this model to analyze the precise sequence of molecular and biological events that lead to the genesis of this disease. On the other hand, the fact that other changes elicited by P210 in the relevant human target cells do not recapitulate features of human CML (eg, enhanced erythroid commitment) underscores the need for continuing improvement in the model, perhaps through a closer approximation of the levels of P210BCR-ABL expression obtained in the more primitive types of CML cells taken directly from patients.

Acknowledgements We thank the members of the Stem Cell Assay Service of the Terry Fox Laboratory for initial cell processing, members of the Flow Cytometry Service for cell sorting, R Van Etten (Center for Blood Research, Boston, MA, USA) for the BCR-ABLDSH2 cDNA, K Humphries and P Lansdorp (both from the Terry Fox Laboratory) for other cDNA probes and antibodies, and StemCell, Novartis, Cangene and Immunex for gifts of recombinant growth factors and other reagents, and R Premji for assistance in manuscript preparation.

References 1 Deininger MWN, Goldman JM, Melo JV. The molecular biology of chronic myeloid leukemia. Blood 2000; 96: 3343–3356. 2 Ben-Neriah Y, Daley GQ, Mes-Masson AM, Witte ON, Baltimore D. The chronic myelogenous leukemia-specific P210 protein is the product of the bcr/abl hybrid gene. Science 1986; 233: 212–214. 3 Konopka JB, Witte ON. Detection of c-abl tyrosine kinase activity in vitro permits direct comparison of normal and altered abl gene products. Mol Cell Biol 1985; 5: 3116–3123. 4 Van Etten RA, Jackson P, Baltimore D. The mouse type IV c-abl gene product is a nuclear protein, and activation of transforming ability is associated with cytoplasmic localization. Cell 1989; 58: 669–678. 5 Daley GQ, Van Etten RA, Baltimore D. Induction of chronic myelogenous leukemia in mice by the P210bcr/abl gene of the Philadelphia chromosome. Science 1990; 247: 824–830. 6 Heisterkamp N, Jenster G, ten Hoeve J, Zovich D, Pattengale PK, Groffen J. Acute leukemia in bcr/abl transgenic mice. Nature 1990; 344: 251–253. 7 Lugo TG, Pendergast AM, Muller AJ, Witte ON. Tyrosine kinase activity and transformation potency of bcr-abl oncogene products. Science 1990; 247: 1079–1082. 8 Evans CA, Owen-Lynch J, Whetton AD, Dive C. Activation of the Abelson tyrosine kinase activity is associated with suppression of apoptosis in hemopoietic cells. Cancer Res 1993; 53: 1735–1738. 9 Druker BJ, Tamura S, Buchdunger E, Ohno S, Segal GM, Fanning S et al. Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 1996; 2: 561–566. 10 Vigneri P, Wang JYJ. Induction of apoptosis in chronic myelogenous leukemia cells through nuclear entrapment of BCR-ABL tyrosine kinase. Nat Med 2001; 7: 228–234. 11 Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, Ford JM et al. Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 2001; 344: 1031–1037. 12 Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C, Gambacorti-Passerini C et al. Hematologic and cytogenetic Leukemia


1012 13

14

15

16

17 18

19 20 21

22

23 24 25

26

27

28

29

30

31

Leukemia

responses to imatinib mesylate in chronic myelogenous leukemia. N Engl J Med 2002; 346: 645–652. Hughes TP, Kaeda J, Branford S, Rudzki Z, Hochhaus A, Hensley ML et al. Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia. N Engl J Med 2003; 349: 1423–1432. Skorski T, Bellacosa A, Nieborowska-Skorska M, Majewski M, Martinez R, Choi JK et al. Transformation of hematopoietic cells by BCR/ABL requires activation of a PI-3k/Akt-dependent pathway. EMBO J 1997; 16: 6151–6161. Roumiantsev S, de Aos IE, Varticovski L, Ilaria RL, Van Etten RA. The src homology 2 domain of Bcr/Abl is required for efficient induction of chronic myeloid leukemia-like disease in mice but not for lymphoid leukemogenesis or activation of phosphatidylinositol 3-kinase. Blood 2001; 97: 4–13. Zhang X, Wong R, Hao SX, Pear WS, Ren R. The SH2 domain of Bcr-Abl is not required to induce a murine myeloproliferative disease; however, SH2 signaling influences disease latency and phenotype. Blood 2001; 97: 277–287. Eaves AC, Eaves CJ. Abnormalities in the erythroid progenitor compartments in patients with chronic myelogenous leukemia (CML). Exp Hematol 1979; 7: 65–75. Strife A, Lambek C, Wisniewski D, Wachter M, Gulati SC, Clarkson BD. Discordant maturation as the primary biological defect in chronic myelogenous leukemia. Cancer Res 1988; 48: 1035–1041. Bedi A, Zehnbauer BA, Barber J, Sharkis S, Jones R. Inhibition of apoptosis by BCR-ABL in chronic myeloid leukemia. Blood 1994; 83: 2038–2044. Maguer-Satta V, Burl S, Liu L, Damen J, Chahine H, Krystal G et al. BCR-ABL accelerates C2-ceramide-induced apoptosis. Oncogene 1998; 16: 237–248. 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. Hariharan IK, Adams JM, Cory S. BCR-ABL oncogene renders myeloid cell line factor independent: potential autocrine mechanism in chronic myeloid leukemia. Oncogene Res 1988; 3: 387–399. Sirard C, Laneuville P, Dick J. Expression of bcr-abl abrogates factor-dependent growth of human hematopoietic M07E cells by an autocrine mechanism. Blood 1994; 83: 1575–1585. Anderson SM, Mladenovic J. The BCR-ABL oncogene requires both kinase activity and src-homology 2 domain to induce cytokine secretion. Blood 1996; 87: 238–244. Zhang X, Ren R. BCR-ABL efficiently induces a myeloproliferative disease and production of excess interleukin-3 and granulocyte– macrophage colony-stimulating factor in mice: a novel model for chronic myelogenous leukemia. Blood 1998; 92: 3829–3840. Li S, Ilaria RL, Million RP, Daley GQ, Van Etten RA. The p190, p210, and p230 forms of the BCR/ABL oncogene induce a similar chronic myeloid leukemia-like syndrome in mice but have different lymphoid leukemogenic activity. J Exp Med 1999; 189: 1399–1412. Jiang X, Ng E, Yip C, Eisterer W, Chalandon Y, Stuible M et al. Primitive interleukin 3 null hematopoietic cells transduced with BCR-ABL show accelerated loss after culture of factor-independence in vitro and leukemogenic activity in vivo. Blood 2002; 100: 3731–3740. Peters DG, Klucher KM, Perlingeiro RC, Dessain SK, Koh EY, Daley GQ. Autocrine and paracrine effects of an ES-cell derived, BCR/ABL-transformed hematopoietic cell line that induces leukemia in mice. Oncogene 2001; 20: 2636–2646. Li S, Gillessen S, Tomasson MH, Dranoff G, Gilliland DG, Van Etten RA. Interleukin 3 and granulocyte–macrophage colonystimulating factor are not required for induction of chronic myeloid leukemia-like myeloproliferative disease in mice by BCR/ABL. Blood 2001; 97: 1442–1450. Zhao RC, Jiang Y, Verfaillie CM. A model of human p210bcr/ABL mediated CML by transducing primary normal human CD34+ cells with a BCR/ABL containing retroviral vector. Blood 2001; 97: 2406–2412. Chalandon Y, Jiang X, Hazelwood G, Loutet S, Conneally E, Eaves A et al. Modulation of p210BCR-ABL activity in transduced primary

32 33 34

35

36 37

38 39

40 41

42 43 44

45

46

47 48

49 50 51

human hematopoietic cells controls lineage reprogramming. Blood 2002; 99: 3197–3204. Kinsella TM, Nolan GP. Episomal vectors rapidly and stably produce high-titer recombinant retrovirus. Hum Gene Ther 1996; 7: 1405–1413. Pear WS, Nolan GP, Scott ML, Baltimore D. Production of high-titer helper-free retroviruses by transient transfection. Proc Natl Acad Sci USA 1993; 90: 8392–8396. Lansdorp PM, Dragowska W. Long-term erythropoiesis from constant numbers of CD34+ cells in serum-free cultures initiated with highly purified progenitor cells from human bone marrow. J Exp Med 1992; 175: 1501–1509. Maguer-Satta V, Petzer AL, Eaves AC, Eaves CJ. BCR-ABL expression in different subpopulations of functionally characterized Ph+ CD34+ cells from patients with chronic myeloid leukemia. Blood 1996; 88: 1796–1804. Ilaria Jr RL, Van Etten RA. The SH2 domain of P210BCR/ABL is not required for the transformation of hematopoietic factor-dependent cells. Blood 1995; 86: 3897–3904. Oda T, Tamura S, Matsuguchi T, Griffin JD, Druker BJ. The SH2 domain of Abl is not required for factor-independent growth induced by Bcr-Abl in a murine myeloid cell line. Leukemia 1995; 9: 295–301. Lozzio BB, Lozzio CB, Bamberger EG, Feliu AS. A multipotential leukemia cell line (K-562) of human origin. Proc Soc Exp Biol Med 1981; 166: 546–550. McWhirter JR, Wang JYJ. An actin-binding function contributes to transformation by the Bcr-Abl oncoprotein of Philadelphia chromosome-positive human leukemias. EMBO J 1993; 12: 1533–1546. Goga A, McLaughlin J, Afar DEH, Saffran DC, Witte ON. Alternative signals to RAS for hematopoietic transformation by the BCR-ABL oncogene. Cell 1995; 82: 981–988. Jiang X, Stuible M, Chalandon Y, Li A, Chan WY, Eisterer W et al. Evidence for a positive role of SHIP in the BCR-ABL-mediated transformation of primitive murine hematopoietic cells and in human chronic myeloid leukemia. Blood 2003; 102: 2976–2984. Issaad C, Vainchenker W. Growth of erythroid colonies in chronic myelogenous leukemia is independent of erythropoietin only in the presence of steel factor. Blood 1994; 84: 3447–3456. Stopka T, Zivny JH, Stopkova P, Prchal JF, Prchal JT. Human hematopoietic progenitors express erythropoietin. Blood 1998; 91: 3766–3772. Ghaffari S, Wu H, Gerlach M, Han Y, Lodish HF, Daley GQ. BCR-ABL and v-SRC tyrosine kinase oncoproteins support normal erythroid development in erythropoietin receptordeficient progenitor cells. Proc Natl Acad Sci USA 1999; 96: 13186–13190. Lopez AF, To LB, Yang YC, Gamble JR, Shannon MF, Burns GF et al. Stimulation of proliferation, differentiation, and function of human cells by primate interleukin 3. Proc Natl Acad Sci USA 1987; 84: 2761–2765. Souza LM, Boone TC, Gabrilove J, Lai PH, Zsebo KM, Murdock DC et al. Recombinant human granulocyte colony-stimulating factor: effects on normal and leukemic myeloid cells. Science 1986; 232: 61–65. Graber SE, Krantz SB. Erythropoietin and the control of red cell production. Annu Rev Med 1978; 29: 51–66. Pharr PN, Ogawa M, Hofbauer A, Longmore GD. Expression of an activated erythropoietin or a colony-stimulating factor 1 receptor by pluripotent progenitors enhances colony formation but does not induce differentiation. Proc Natl Acad Sci USA 1994; 91: 7482–7486. Socolovsky M, Lodish HF, Daley GQ. Control of hematopoietic differentiation: lack of specificity in signaling by cytokine receptors. Proc Natl Acad Sci USA 1998; 95: 6573–6575. Metcalf D. Lineage commitment and maturation in hematopoietic cells: the case for extrinsic regulation. Blood 1998; 92: 345–348. Hogge DE, Lansdorp PM, Reid D, Gerhard B, Eaves CJ. Enhanced detection, maintenance and differentiation of primitive human hematopoietic cells in cultures containing murine fibroblasts engineered to produce human Steel factor, interleukin-3 and granulocyte colony-stimulating factor. Blood 1996; 88: 3765–3773.