Cytotoxic T Cells Overcome BCR-ABL-mediated Resistance to Apoptosis1

Chronic myeloid leukemia is a disease marked by expanded clonal hematopoiesis; it is incurable by chemotherapy or radiation but is cured in a majority of patients receiving bone marrow transplantation from nonidentical sibling donors, an outcome generally attributed to a T cellmediated graft-versus-leukemia effect. In this report, we examine the effect of the P210Bl R'ABI fusion protein of the BCR-ABL oncogene, the molecular hallmark of chronic myelogenous leukemia, on the sensitivity of mouse cell lines to apoptosis induced by chemotherapy, radiation, or activated cytotoxic T lymphocytes (CTLs). We find that, although cells expressing P210B< R-ABI-by gene transfer are more resistant than their normal counterparts to apoptosis induced by chemotherapy or radiation, they are equally susceptible to apoptosis induced by alloreactive CTLs. These results show that CTLs overcome BCR-ABL-mediated resistance to apoptosis and, therefore, provide a biological correlation for the suc cess of allogeneic bone marrow transplantation in chronic myelogenous leukemia. Introduction CML3 is a hematological malignancy characterized by an initial chronic phase of expanded clonal hematopoiesis with continued dif ferentiation into mature granulocytes; this phase invariably progresses to blast crisis, a terminal stage resembling acute leukemia. The cytogenetic hallmark of CML, the Philadelphia chromosome (1), results from the BCR-ABL gene rearrangement, which encodes a Mr 210,000 fusion protein (P210BCR"ABL)with cytoplasmic tyrosine kinase activ ity. The BCR-ABL rearrangement likely initiates CML, because mice acquire a disease resembling chronic leukemia when transplanted with bone marrow cells that are infected with a retrovirus encoding BCRABL (2). A crucial effect of BCR-ABL expression in CML is the prolongation of hematopoietic cell survival by inhibition of apoptosis, the physiological form of autonomous cell death. We and others have shown that p210BCR"ABLconfers upon hematopoietic cells the ability to survive treatments, such as growth factor deprivation or genotoxic chemotherapy, that induce apoptosis in normal cells (3, 4). These findings probably account for the inability of standard chemotherapy to cure patients with CML. Nonetheless, CML is curable by myeloablative therapy followed by bone marrow transplantation. Relapse of CML occurs in <10% of recipients of marrow from a nonidentical sibling donor, but the relapse rate increases to 50% when the marrow is depleted of T cells or is from an identical twin donor (5). These findings strongly suggest that allogeneic donor T cells exert an immunological antitumor effect in CML. Because CTLs, the major effectors of antitumor immunity, are capable of inducing both apoptosis and osmotic lysis in their targets Received 11/4/94; accepted 12/16/94. 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. 1Supported by Grant CA-15396 from the National Cancer Institute. R. J. J. is a Leukemia Society of America Scholar. 2 To whom requests for reprints should be addressed, at Division of Bone Marrow Transplantation, The Johns Hopkins Oncology Center, 6(X)North Wolfe Street, Baltimore, MD 21287-8985. 3 The abbreviations used are: CML, chronic myelogenous leukemia; CTL, cytotoxic T lymphocyte; IL-3, interleukin 3. (6), we were interested to test the effect of p210BCR~ABLexpression in targets of CTL-mediated killing. Interestingly, we find that, although P2JQBCR-ABLproteC(S ceiis from chemotherapy-induced apoptosis, it fails to protect cells from apoptosis induced by activated CTLs. These findings provide a biological explanation for the relative efficacies of various treatments of CML and provide a rationale for the use of immunotherapy in cancers that are resistant to chemotherapy. Materials and Methods Cell Lines. Studies were performed on two IL-3-dependcnt mouse cell lines, Ba-F3 (early B-lineage; Ref. 7) and FDC-P1 (myeloid; Ref. 8). The cell lines, induced to constitutively express P210I1CR~ABLby rctroviral infection with a full-length BCR-ABL cDNA sequence, were provided by R. A. van Etten and D. Baltimore (Ba-F3p:ln; Ref. 9) and S. Cory (FDC-P1P21(1; Ref. 10). The cell lines were maintained in complete (containing vitamins, amino acids, and lipids) serum-free medium (Serum-Free and Protein-Free Hybridoma Medium; Sigma Chemical Co., St. Louis, MO) in 5% CO2 at 37°Cwith 1(K) units recombinant mouse IL-3 (or 10% conditioned medium of the WEHI-3 cell line as a source of IL-3) added to the parental cell lines. Induction and Assays of Activated CTLs. Spleen cells (4 x 10'') from C57BL/10 mice (BIO; H-2h) were cultured with 2 x 10" irradiated (3000 cGy) BALB/c (H-2d) or C3H/HeJ (H-2k) spleen stimulators in 2 ml EHAA medium (Biofluids, Rockville, MD) containing 10% PCS, 5 X 10~5 M 2-mercaptoethanol, glutamine, and antibiotics (complete medium). After 5 days of culture, viable CTLs were isolated by centrifugation over Ficoll (Pharmacia, Uppsala. Sweden), counted, and tested for induction of DNA fragmentation or osmotic lysis of 'Hor MCr-labeled targets, respectively. DNA fragmentation was assessed at various effector CTL:target (E:T) ratios by the JAM test (11). Osmotic lysis was assessed by plating CTL at various E:T ratios with 1()4 labeled targets in a standard chromium release assay (II). Analysis of Cellular Viability following Exposure to Activated CTLs or Cytotoxic Agents. Ba-F3 or Ba-F3'"'" cells were treated with graded con centrations of VP-16 (0-80 fig/ml for 30 min), were irradiated (0-2000 cGy), or were cultured for 2 h with BIO CTL generated against BALB/c stimulators as described above. The viability of cells exposed to VP-16, irradiation, or CTLs was determined by methylcellulose clonogenic assays as described previously (4). DNA Fragmentation Assays. FDC-P1 or FDC-P1P21" cells were incu bated in medium alone or with activated CTLs for 2 h. Total genomic DNA was then isolated after SDS lysis and proteinase K digestion as described (12). Oligonucleosomal DNA fragments were separated by agarose gel (2%) electrophoresis and visualized by staining with ethidium bromide. One hundred base-pair size markers (GIBCO-BRL, Gaithersburg, MD) were also separated by electrophoresis as controls. In Situ Nick-end Labeling of Apoptotic Nuclei. FDC-P1P21°cells were incubated with activated CTLs for 2 h and then fixed in cold 4% buffered formaldehyde (pH 7.4), followed by cold 70% ethanol. The fixed cells were then incubated with 20 /¿g/mlproteinase K (Sigma), washed in cold PBS, and immersed in 100 /¿Ireaction buffer [0.2 M potassium cacodylate, 25 rnw Tris-HCl (pH 6.6), 0.25 mg/ml BSA, and 2.5 mM cobalt chloride) supple mented with terminal deoxynucleotidyl transferase (0.3 e.u./fil) and biotinylated dUTP. Each experiment was performed with a negative control (without b-dUTP) and a positive control [10-min pretreatment with 1 ng/ml DNase (Sigma), dissolved in reaction buffer). Following incubation at 37°Cfor 30 min, the cells were washed in PBS and incubated with 100 /il FITC-avidin solution [4X saline-sodium citrate buffer (Sigma), 2.5 fig/ml fluoresceinated avidin, 0.1% Triton X-100, and 5% w/v nonfat dry milk) at room temperature for 30 min in the dark. The cells were washed in PBS with 0.1% Triton X-100