Selective induction of apoptosis in leukemic B-lymphoid cells by a CD19-specific TRAIL fusion protein

Abstract

Although the treatment outcome of lymphoid malignancies has improved in recent years by the introduction of transplantation and antibody-based therapeutics, relapse remains a major problem. Therefore, new therapeutic options are urgently needed. One promising approach is the selective activation of apoptosis in tumor cells by the tumor necrosis factor-related apoptosis-inducing ligand (TRAIL). This study investigated the pro-apoptotic potential of a novel TRAIL fusion protein designated scFvCD19:sTRAIL, consisting of a CD19-specific single-chain Fv antibody fragment (scFv) fused to the soluble extracellular domain of TRAIL (sTRAIL). Potent apoptosis was induced by scFvCD19:sTRAIL in several CD19-positive tumor cell lines, whereas normal blood cells remained unaffected. In mixed culture experiments, selective binding of scFvCD19:sTRAIL to CD19-positive cells resulted in strong induction of apoptosis in CD19-negative bystander tumor cells. Simultaneous treatment of CD19-positive cell lines with scFvCD19:sTRAIL and valproic acid (VPA) or Cyclosporin A induced strongly synergistic apoptosis. Treatment of patient-derived acute B-lymphoblastic leukemia (B-ALL) and chronic B-lymphocytic leukemia (B-CLL) cells resulted in strong tumoricidal activity that was further enhanced by combination with VPA. In addition, scFvCD19:sTRAIL prevented engraftment of human Nalm-6 cells in xenotransplanted NOD/Scid mice. The pre-clinical data presented here warrant further investigation of scFvCD19:sTRAIL as a potential new therapeutic agent for CD19-positive B-lineage malignancies.

Fig. 1

CD19-restricted binding and serum stability of scFvCD19 and scFvCD19:sTRAIL. a Scheme of homotrimeric scFvCD19:sTRAIL consisting of the N-terminal (N-) hemagglutinin-tag (HA) genetically fused to scFvCD19 and C-terminal (C-) sTRAIL. The intrinsic trimerization domain of sTRAIL leads to formation of active homotrimers. b CD19-positive SEM cells were stained with scFvCD19 or scFvCD19:sTRAIL (dark grey), respectively. PBS or CD CHO medium served as control for background staining (light grey). Binding was assessed by flow cytometry. Bold black lines blocked binding after pre-incubation with tenfold molar excess of parental antibody. c and d scFvCD19 and scFvCD19:sTRAIL were kept in human serum at 37°C for the indicated time intervals. c Samples were analyzed by flow cytometry and residual binding to SEM cells was calculated from 4 independent experiments. Values are normalized to binding of freshly prepared samples. d CD19-positive Reh cells were treated with subsaturating concentrations of scFvCD19:sTRAIL after incubation in human serum for the indicated length of time. Cell death was measured by AnnexinV/PI staining after 96 h. Experiments were performed 4 times independently. Percent cell death was calculated relative to maximum apoptosis obtained with freshly prepared samples. Error bars standard error of the mean (SEM)

Fig. 2

Induction of CD19-restricted apoptosis by scFvCD19:sTRAIL in leukemia-derived cell lines, but not in MNCs from healthy donors. a Various CD19-positive cell lines (Reh, Nalm-6 and SEM) and the CD19-negative cell line CEM were treated with increasing concentrations of scFvCD19:sTRAIL. After treatment for 96 h, cell death was measured by hypotonic PI staining of nuclei. Data are mean values ± SEM from 3 independent experiments. b Reh, Nalm-6 and CEM control cells were treated with scFvCD19:sTRAIL (100 ng/ml) in the absence or presence of 100-fold molar excess of parental CD19 antibody N6B1, tenfold molar excess of TRAIL neutralizing antibody 2E5, or the respective isotype antibodies. (+): treatment with scFvCD19:sTRAIL plus the respective antibody. Cell death was measured 96 h after treatment by hypotonic PI staining. Values shown are mean values + SEM from 3 independent experiments. c Reh, Nalm-6 and CEM cells were treated with 10 ng/ml of scFvCD19:sTRAIL or CD CHO medium. After treatment for 24 h, exposure of phosphatidylserine to the outer cell membrane was assessed by AnnexinV/PI staining. Percentages in the lower right quadrant indicate early apoptotic cells. d Cells were treated with 100 ng/ml of scFvCD19:sTRAIL (+) or medium control (−) for 48 h. Characteristic indicators of apoptosis were revealed by immunoblotting: PARP and its cleavage product (arrow) and disappearance of pro-caspase 3 with actin as a loading control. e Isolated MNCs from healthy donors were treated with 100 ng/ml of scFvCD19:sTRAIL, medium control, 8 μg/ml Actinomycin D or 300 μg/ml VPA for 24 h. Apoptosis was assessed by loss of ΔΨ and values are mean values + SEM from 3 independent experiments

Fig. 3

Induction of apoptosis in CD19-negative bystander tumor cells. Mixed culture experiments with CD19-positive Ramos cells a and CD19-negative Jurkat cells b were performed at varying T:B ratios. Cells were treated with 100 ng/ml scFvCD19:sTRAIL in the presence or absence of the TRAIL neutralizing antibody 2E5. The differential fluorescent labeling of target and bystander cell populations was used to separately evaluate apoptosis induction by ΔΨ after 24 h of treatment. Values given are mean values + SEM from 3 independent experiments

Fig. 4

Synergistic tumoricidal effect of scFvCD19:sTRAIL in combination with CsA or VPA for various B-ALL cell lines. Treatment of Reh, Nalm-6, SEM and CEM control cells with antileukemic drugs CsA (a) or VPA (b), respectively, was performed in combination with scFvCD19:sTRAIL (10 ng/ml). For VPA treatment, combination of VPA plus scFv425:sTRAIL was included as an additional control. Final concentrations of 4 μg/ml for Reh, Nalm-6 and CEM cells or 10 μg/ml of CsA for SEM cells and 100 μg/ml for Reh, SEM and CEM cells or 150 μg/ml of VPA for Nalm-6 cells, respectively, were applied. Cell death was determined by AnnexinV/PI staining after 24 h. Specific cell death was defined as percentage of dead cells over background of spontaneous cell death in untreated control samples. C-Index values were calculated as described in “Materials and methods” and synergism (C-Index < 1.0) is indicated by (*). Values given are mean values + SEM from 3 independent experiments. c After combination treatment with VPA, expression levels of activating TRAIL-R1/2 were monitored by flow cytometry for the cell lines investigated here. Cells were stained with TRAIL-R1- (left panel) or TRAIL-R2-specific antibody (right panel), respectively. Murine IgG1 antibody served as control for background staining (light grey). Binding of receptor-specific antibody in untreated samples is depicted in dark grey. Black lines represent receptor expression of VPA-treated samples, dotted lines of scFvCD19:sTRAIL treated samples. Receptor expression of combination-treated samples is depicted in bold black lines

Fig. 5

Induction of cell death in B-ALL-derived MNCs by scFvCD19:sTRAIL, and synergistic effects on primary B-CLL cells by co-treatment with VPA. a MNCs from pediatric B-ALL patients were treated with 100 ng/ml of scFvCD19:sTRAIL or scFv425:sTRAIL, respectively. Specific cell death was determined after 24 h by AnnexinV/PI staining. Several B-CLL samples sensitive for scFvCD19:sTRAIL (b) and the only resistant sample (c) were treated for 24 h with 300 μg/ml of VPA in the presence or absence of scFvCD19:sTRAIL. Specific cell death was determined using AnnexinV/PI staining. C-Index values were calculated as described in “Materials and methods”; synergism is indicated by (*). All experiments were performed in triplicates and values given are mean values ± SEM

Fig. 6

Inhibition of engraftment in a murine model of xenotransplanted human B-ALL cells by scFvCD19:sTRAIL. Female NOD/Scid mice were injected intravenously into the lateral tail vein with 10⁶ Nalm-6 cells on day 0. On days 3,5 and 7 mice were treated with 4 μg of scFvCD19:sTRAIL per dose intravenously, or with scFv425:sTRAIL, and one group received medium as a control. Mice were monitored for 112 days. At first signs of paralysis of hindlegs and/or loss of > 20% of body weight, mice were euthanized. Percent survival was calculated for 9 mice from 2 independent experiments, which were treated with medium, scFv425:sTRAIL and scFvCD19:sTRAIL, respectively. P values were calculated using log rank test and the median survival was calculated for each group