The genes encoding the AML1 (RUNX1) or CBFβ subunits of core binding factor (CBF) are commonly altered by translocation or mutation in human leukemias. Because CBF oncoproteins slow G1, we sought to determine whether mutations that accelerate G1 potentiate their ability to induce transformation. Wild-type or p16INK4ap19ARF (−/−) marrow cells transduced with CBFβ-smooth muscle myosin heavy chain (SMMHC) were transplanted into wild-type, syngeneic recipients. CBFβ-SMMHC significantly increased the development of acute leukemias from marrow lacking the overlapping p16p19 genes, based on analysis of Kaplan-Meier event-time distributions. Wild-type marrow was also transduced with vectors expressing either E7 alone or both E7 and CBFβ-SMMHC. Combining oncogenes again increased leukemia formation. Exposing mice transplanted with CBFβ-SMMHC-transduced cells to a mutagen, ethylnitrosourea, markedly accelerated leukemogenesis compared to expressing CBFβ-SMMHC with loss of p16p19, indicating the need for multiple “hits” for transformation. The INV/p16p19 and INV/E7 leukemias were lymphoid and were clonal and retransplantable. Overall, these findings indicate that CBF mutations cooperate with genetic alterations that accelerate G1 to induce acute leukemia.
One or both alleles of the AML13 (RUNX1) subunit of CBF harbor somatic point mutations in 5% of patients with acute myeloid leukemia. Subsets of acute leukemias also express fusion oncoproteins that inhibit CBF: 12% of AML cases harbor t(8;21)(q22;q22), which encodes AML1-ETO; 25% of pediatric B-precursor ALL (B-ALL) cases express TEL-AML1 from the t(12;21)(p13;q22) chromosome; and the CBFβ subunit of CBF is linked to the rod domain of SMMHC by inv16(p13;q22) or t(16;16)(p13;q22) in 10% of AML patients. CBFβ-SMMHC inhibits CBF DNA binding by sequestering AML1 and other CBFα subunits in multimers that form via the SMMHC domain, and AML1-ETO and TEL-AML1 each repress the transcription of genes normally activated by CBF. Mice expressing CBFβ-SMMHC or AML1-ETO and those lacking AML1 or CBFβ each fail to develop definitive hematopoiesis, demonstrating that these CBF oncoproteins inhibit CBF activities in vivo (1).
Exposure of chimeric mice expressing CBFβ-SMMHC to one dose of a mutagen, ENU, results in the majority developing AML by 5 months of age (2). Similarly, mice expressing AML1-ETO in hematopoietic cells develop AML, and less commonly T-ALL, only when exposed to ENU (3, 4). Some AML patients in long-term remission retain the AML1-ETO fusion gene in a small fraction of their marrow cells, suggesting that AML1-ETO is present in potential leukemic precursor cells lacking the additional mutagenic “hits” necessary for transformation (5). Thus, it appears that CBF oncoproteins cannot, alone, transform hematopoietic stem cells but instead cooperate with additional mutations to induce acute leukemias in mice and humans.
Inhibition of CBF blocks myeloid and lymphoid differentiation in normal marrow, prevents MPO induction in 32D cl3 cells, and slows apoptosis of Ba/F3 cells in response to DNA damage (2, 6, 7). In addition, inhibition of CBF slows G1 progression in 32D cl3, Ba/F3, or U937 cells, and overexpression of AML1 in 32D cl3 cells accelerates G1 (6, 8, 9, 10, 11). Exogenous cdk4 or cyclin D2 overcomes delayed cell cycle progression of 32D cl3 and Ba/F3 cells resulting from blockade of CBF trans-activation (6), and c-Myc acts similarly.4 Taken together, these findings suggest that hematopoietic cells expressing CBFβ-SMMHC or alterations of AML1 might be blocked at an early stage of differentiation but are incapable of proliferation. We therefore sought to determine whether acceleration of G1 would cooperate with CBFβ-SMMHC to induce acute leukemia in vivo.
Materials and Methods
The MING retroviral vector encodes NGFR-GFP, residues 1–277 of the human low-affinity NGFR (p75) containing its extracellular and trans-membrane domain, linked to the NH2 terminus of full-length enhanced GFP, residues 1–238. The HPV16-E7 or CBFβ-SMMHC cDNAs were subcloned upstream of the IRES to generate MING-E7 or MING-INV. The E7 cDNA was also cloned downstream of the IRES in MING-INV, deleting the majority of the NGFR-GFP cDNA, to produce MING-INV/E7.
Preparation of Packaging Lines.
Ecotropic φCRE packaging cells (12) were cultured in DMEM with 10% heat-inactivated calf serum. They were transfected by calcium phosphate precipitation with 20 μg of MING, MING-INV, or MING-E7 and 2 μg of pSV40-puro. Pooled transfectants were selected using 2 μg/ml puromycin. Cells (106) were then immunoaffinity purified with biotin-conjugated anti-NGFR monoclonal antibody (American Type Culture Collection; HB8737) and streptavidin-coated magnetic beads (Stem Cell Technologies, Vancouver, British Columbia, Canada). MING-INV/E7 transductants were subcloned by limiting dilution. The optimum producer among 50 subclones was identified using viral RNA slot blot analysis. In brief, viral particles were precipitated using 10% polyethylene glycol 8000 in 0.5 m NaCl. RNA was then extracted using phenol and loaded onto a nylon filter using a slot blot apparatus. For comparing the four packaging lines, a 1.7-kb SphI/EcoRI fragment containing the 5′ LTR and adjacent retroviral sequences was used as probe using a Northern blot procedure (6).
Transduction and Transplantation of Murine Bone Marrow.
Donor C57BL/6 or C57BL/6 p16INK4a/p19ARF (−/−) mice (13) were treated with 150 mg/kg 5-fluorouracil i.v., and 3 days later marrow cells were flushed from the femurs and tibias using QBSF-58 medium (generously provided by Quality Biological, Inc.). After red cell lysis with ammonium chloride, the cells were resuspended in QBSF-58, supplemented with 2 mm glutamine, 100 ng/ml murine interleukin 6, 100 ng/ml murine interleukin 3, and 50 ng/ml murine SCF (PeproTech) and cultured for 48 h at 37°C, 5% CO2. Confluent φCRE packaging lines were irradiated to 40 Gy before coculture with 8 × 106 marrow cells/100-mm dish with 4 μg/ml Polybrene for an additional 2 days. Transduced marrow cells (2 × 106) were then injected into the tail veins of recipient C57BL/6 mice that had received 6.0 Gy of γ-irradiation the day before and 3.5 Gy the day of injection. ENU was used at 50 mg/kg i.p. on day 35 after transplant when indicated. Leukemia transplantation to secondary recipients was by i.p. injection.
FACS Analysis of Leukemic Cells.
Murine marrow, thymus, or spleen cells (106) were stained with biotin-conjugated antibodies at 1 μg/106 cells for 30 min at 4°C, washed, incubated with streptavidin-conjugated R-phycoerythrin (BD PharMingen) for 30 min at 4°C, and fixed with 1.0% paraformaldehyde, 0.01 m HEPES (pH 7.4). Antibodies used recognized Sca-1 (Ly6A/E, stem cell), TCRβ (H57–597, T-lineage), B-220 (RA3–6B2, B-lineage), Mac-1 (M1/70, macrophage), Gr-1 (RB6–8C5, granulocyte), Ter-119 (erythroid), CD4 (GK1.5, T-lineage), and CD8α (53–6.7, T-lineage; BD PharMingen, San Diego, CA). TCR-Vβ subtypes were determined using a screening antibody panel kit (BD PharMingen). Immunophenotypes were analyzed with a FACScan flow cytometer (Becton Dickinson).
The end point of this study was time to leukemia. Event time distributions were estimated by the method of Kaplan and Meier. Hazard ratios were obtained using the Cox model. When the Cox model did not converge, the log-rank P is presented. Statistical computations were performed using the SAS or EGRET PC packages (14, 15).
Development of Retroviral Vectors.
The control vector (MING), expressing only NGFR-GFP, and the vectors encoding E7, CBFβ-SMMHC (designated INV for brevity), or INV and E7 are diagrammed in Fig. 1,A. These vectors were packaged using φCRE cells, and the viral titers of the resulting four producer populations were then compared (Fig. 1,B). The titers of the INV and E7 vectors were similar. The INV/E7 vector had an ∼3-fold lower titer, and the MING vector had a 2-fold higher titer. These relative titers are consistent with the relative transduction of Sca-1+ marrow cells by the three packaging lines expressing NGFR-GFP (Fig. 1 C). Sca-1 is a marker of stem/progenitor cells. The MING, INV, and E7 vectors transferred GFP expression to 27, 10, and 14% of Sca-1+ marrow cells, respectively.
Retrovirally Expressed INV Cooperates with ENU to Produce Leukemia.
To validate our strategy of retroviral marrow transduction for obtaining INV-dependent leukemias, we treated mice transplanted with INV-transduced cells with a single dose of ENU, a mutagen. As with untreated chimeric INV mice (2), INV alone induced only a low incidence of leukemia and not until after 1 year of age (Fig. 2,A). One of these leukemias was a T-ALL, and the other was a Gr-1+, Mac-1+ AML with prominent primary granules. In contrast and as with chimeric INV mice, the combination of INV and ENU markedly accelerated leukemogenesis compared with ENU alone (Fig. 2 B; hazard ratio, 8.0; 95% confidence interval, 1.9–33; P = 0.0005). Each of the ENU and INV/ENU leukemias stained for TCRβ, but not for B lymphoid or myeloid markers, and were therefore classified as T-ALL. All mice with T-ALL had grossly enlarged thymuses and spleens, and these organs as well as the bone marrow were largely replaced with leukemic blasts, as assessed by morphological examination of Wright’s-Giemsa-stained cytospins (not shown).
Loss of p16p19 Cooperates with INV to Induce Acute Leukemia.
The MING or INV vectors were introduced into marrow cells isolated from C57BL/6 mice lacking the overlapping p16INK4a and p19ARF genes, and the transduced cells were then injected into irradiated, wild-type, syngeneic recipients. p16INK4a is a cyclin-dependent kinase inhibitor that predominantly inhibits cdk4 and cdk6. Lack of p19ARF, the murine homologue of human p14ARF, is reported to increase Mdm2 and thereby decrease p53 activity. p53 can slow G1 or induce apoptosis in hematopoietic cells (16). Leukemia incidence for these two cohorts is shown in Fig. 2 C. Because the MING/p16p19 group had no leukemias, its hazard ratio versus the INV/p16p19 cohort could not be calculated, but the log-rank test shows a significant difference between these groups (P = 0.04). Two INV/p16p19 leukemias were T-ALL (at 5 and 6.5 months), two were B-ALL (at 8 and 11 months), and one was an AML (at 14 months). The AML had a morphology similar to FAB M4 AML, with both monoblasts and granulated myeloblasts, but no malignant eosinophils.
E7 Cooperates with INV to Induce Acute Leukemia.
The E7 protein encoded by human papillomavirus type 16 accelerates the G1 to S cell cycle transition by inactivating pRb (17). Marrow cells from C57BL/6 mice were transduced with either the E7 or the INV/E7 vector and then transplanted into syngeneic recipients. Leukemia-free survival data for a 20-month period is presented in Fig. 2 D. Leukemia developed more rapidly and at a higher incidence in the INV/E7 cohort (hazard ratio, 3.2; 95% confidence interval, 0.7–18; P = 0.14). Although this difference did not reach statistical significance, the difference in leukemia-free survival between INV/E7 mice and E7 or INV mice is likely biologically meaningful, especially given the lower titer of the INV/E7 vector. All eight of the INV/E7 leukemias were ALL. One of the E7 leukemias was T-ALL, and the other was B-ALL. Of those analyzed, three of the INV/E7 T-ALLs were CD4+/CD8+, one was CD4−/CD8+, and one was CD4+/CD8−.
INV-induced Leukemias Were Clonal and Retransplantable.
TCR-Vβ subset analysis of two INV/p16p19, two INV/E7, and one E7 T-ALL demonstrated that each was clonal, with TCR-Vβ-6 (two leukemias), Vβ-11, Vβ-13, or Vβ-14 rearrangements, respectively. Representative data are shown in Fig. 3. Each of the T-ALLs bound the CD4/CD8 antibody mixture (Y axis), and TCR-Vβ-6 was expressed by the T-ALL studied in the experiment shown (X axis). PCR identified the retroviral vector in the leukemias analyzed (not shown). Cells (2 × 106) from two INV/p16p19 leukemias (one T-ALL and one B-ALL) and from four of the INV/E7 T-ALLs were injected into syngeneic, nonirradiated secondary recipients, and all gave rise to leukemias of the same morphology and phenotype after 1–2 months (not shown).
Using two different approaches, expression in p16INK4a/p19ARF (−/−) marrow cells or coexpression with E7, we have provided evidence indicating that acceleration of G1 cooperates with CBFβ-SMMHC to induce acute leukemia in mice. This is the first experimental demonstration of cooperative transformation by two functional alterations frequently present in acute leukemias, CBF inhibition and G1 stimulation. Our findings support a model in which mutations that accelerate G1 potentiate the ability of CBF oncoproteins or AML1 mutations to contribute to leukemogenesis (Fig. 4). For example, genetic alterations that accelerate G1 might allow myeloid stem cells to proliferate despite the presence of CBFβ-SMMHC. G1 stimulation might even enable increased levels of CBFβ-SMMHC, which would in turn more effectively block differentiation or apoptosis.
Our findings also suggest that even the two genetic alterations we provided in the INV/E7 or INV/p16p19 experimental groups were insufficient for transformation. Leukemias were not obtained in all of the transplanted mice, and a single cell/mouse gave rise to the leukemias observed. If expression of CBFβ-SMMHC with E7 or in p16INK4a/p19ARF (−/−) marrow cells was sufficient for transformation, then multiple (polyclonal) leukemias might have well developed in each transplanted mouse. In addition, combining INV with ENU significantly increased the rate of leukemia development compared with INV/p16p19 (P = 0.004) or INV/E7 (P = 0.005). Thus, ENU may both induce a mutation that accelerates G1 and provide additional mutations that are rate-limiting for transformation by the INV/E7 or INV/p16p19 combinations. In the future, we will analyze INV/ENU leukemias for mutations in p16INK4a/p19ARF, p53, Rb, and other proteins that affect G1 progression. Also, because mice lacking either p16INK4a or p19ARF individually are now available (18, 19), assessment of the ability of INV to transform their marrow progenitors will clarify whether INV cooperated with alteration of the Rb or p53 pathways in our INV/p16p19 cohort.
Most of the leukemias obtained in these experiments were ALL, mainly T-ALL, rather than AML, regardless of whether CBFβ-SMMHC was coexpressed with E7, lack of p16 and p19, or random mutations induced by ENU. Perhaps the murine stem cell virus retroviral promoter expresses more efficiently in lymphoid than in myeloid progenitors. Alternatively, perhaps E7, loss of p16p19, or ENU selected for lymphoid leukemias. Hypermethylation inactivates the promoter of the gene encoding p15INK4b, a cdk inhibitor, in 60–80% of AML cases, including patients with CBF leukemias (20). Overexpression of c-Myc from double minute chromosomes also occurs in AML (21). In the future, we will determine whether expressing CBFβ-SMMHC with c-Myc or loss of p15 might more effectively generate AML. Higher titer vectors or use of a different growth factor mixture might enable transduction of a greater number of myeloid progenitors, which in turn might also favor the formation of AML. Nevertheless, the involvement of CBF inhibition in a subset of human lymphoid leukemias and the involvement of CBF in regulating the normal proliferation and differentiation of both myeloid and lymphoid cells suggest that our findings are relevant to the generation of CBF leukemias in general.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Suported by NIH Grant HL51388 and a grant from the Children’s Cancer Foundation (to A. D. F.), by a grant from the National Foundation for Cancer Research (to C. I. C. and A. D. F.), and by NIH Training Grant T32 CA60441 (to Y. Y.). A. D. F. is a Scholar of the Leukemia and Lymphoma Society.
The abbreviations used are: AML, acute myelogenous leukemia; ALL, acute lymphocytic leukemia; T-ALL, T-lineage ALL; B-ALL, B-lineage ALL; CBF, core binding factor; SMMHC, smooth muscle myosin heavy chain; NGFR, nerve growth factor receptor; GFP, green fluorescent protein; IRES, internal ribosome entry site; ENU, ethylnitrosourea; LTR, long terminal repeat; TCR, T-cell receptor; cdk, cyclin-dependent kinase.
F. Bernardin, Y. Yang, C. I. Civin, and A. D. Friedman, unpublished data.
We thank D. Bodine for advice regarding marrow transduction, M. Britos-Bray for technical assistance, V. Tanavde for help with fluorescence-activate cell sorter analysis, P. Liu for the CBFβ-SMMHC cDNA, K. Cho for the E7 cDNA, and N. E. Sharpless and R. A. DePinho for the p16INK4a/p19ARF (−/−) mice.