BH3-interacting domain death agonist (BID) is a crucial element in death signaling pathways and is recognized as an intracellular link connecting the intrinsic mitochondrial apoptotic and extrinsic death receptor–mediated apoptotic pathways. Herein, we describe experiments conducted with a fusion protein, which was generated by fusing a human epidermal growth factor receptor-2 (HER2)–specific single-chain antibody with domain II of Pseudomonas exotoxin A and the truncated active BID (tBID). These experiments extend our previous work on several other immuno-proapoptotic proteins. Specifically, by excluding cells with undetectable HER2, we showed that the secreted immuno-tBID molecule selectively recognized and killed HER2-overexpressing tumor cells in vitro by attacking their mitochondria and inducing their apoptotic death. This apoptosis could only be inhibited partially by caspase pan-inhibitor zVAD and mitochondrial protector TAT-BH4. Subsequently, we transferred the immuno-tbid gene into BALB/c athymic mice bearing HER2-positive tumors together with other immuno-proapoptotic proteins using i.m. injections of liposome-encapsulated vectors. The expression of the immuno-tbid gene suppressed tumor growth and prolonged animal survival significantly. We also shortened the translocation domain of Pseudomonas exotoxin A II to only 10–amino acid sequence, which were crucial for furin cleavage. The new recombinant molecule retained the translocation efficiency and the ability of specific killing HER2-positive tumor cells. Our data showed that, compared with the toxins employed before, the chimeric immuno-tBID molecule can not only specifically recognize HER2-positive tumor cells but also certainly induce apoptosis even in the presence of zVAD and TAT-BH4, thereby suggesting an alternative approach to treating HER2/neu-positive tumors. [Mol Cancer Ther 2008;7(7):1890–9]

Tumors are thought to arise through of a series of mutations that alter the appropriate functions of oncoproteins. A large number of these oncoproteins are transmembrane molecules that participate in signal transduction. The human epidermal growth factor receptor-2 (HER2), a member of the epithelial growth factor receptor family, transduces cell signaling and plays key roles in cell differentiation, adhesion, and motility (1). It is expressed highly in many cancer types (26) but lowly in most normal adult tissues. From experimental and clinical viewpoints, HER2 is a key player in cell communication, whose overexpression may result in poor cancer prognoses, and was therefore chosen as a target for current clinical trials.

Two basic pathways, extrinsic and intrinsic, can direct a cell to undergo apoptosis. The extrinsic apoptotic pathway is mediated by the binding of apoptosis-inducing ligands, including the tumor necrosis factor, tumor necrosis factor–related apoptosis-inducing ligand, and the Fas ligand to their receptors (tumor necrosis factor receptor and Fas) on the cell surface. Alternatively, the intrinsic pathway is regulated by cytochrome c release (7) and is induced typically by stress, radiation, and chemotherapeutic drugs. Each of these pathways converges in the activation of cytosolic aspartate-specific proteases (caspases) that degrade cellular proteins and commit the cell to self-destruction.

BH3-interacting domain death agonist (BID) is a proapoptotic BH3 domain–only member of the Bcl-2 family thought to exert its death-inducing effect by stimulating the release of proapoptotic factors [cytochrome c, apoptosis-inducing factor (AIF), and procaspase-9] from the mitochondria. Full-length BID, normally present in the cytosol, is cleaved by activated caspase-8 into a 15.5-kDa COOH-terminal fragment (tBID), which translocates to the mitochondria and causes cytochrome c release (8, 9). Because BID is recognized as an intracellular link connecting the extrinsic and intrinsic apoptotic pathways, the application of BID to kill tumor cells may be an efficient strategy for treating malignant disease.

During the past decade, we were devoted to studying sensitive, efficient, and specific cancer gene therapies, especially with regard to cancers overexpressing HER2. Several immuno-proapoptotic proteins (immuno-caspase-3, immuno-caspase-6, and immuno-granzyme B) that comprise a HER2-specific single-chain antibody (e23sFv; ref. 10), translocation domain of Pseudomonas exotoxin A (PEA; refs. 11, 12), which can introduce COOH-terminal domain transferring from endosome to cytosol, and proapoptotic proteins were shown to have potent selective antitumor effects both in vitro and in vivo (1317). However, the immuno-proapoptotic proteins could be inhibited or even blocked by some intracellular caspase inhibitors and/or mitochondrial protectors, and as a heterogeneous protein, PEA II employed could also induce the failed treatment.

To develop a more sensitive strategy for applying antitumor therapy, we generated a chimeric molecule, which replaced the domain III in immuno-proapoptotic proteins with tBID to generate immuno-tBID (e23sFv-PEA II-BIDΔ1-60/75). The specific cell-killing efficiency of e23sFv-PEA II-BIDΔ1-60/75 was conducted with and without the presence of caspase pan-inhibitor zVAD and mitochondrial protector TAT-BH4. To generate a recombinant therapeutic gene without the presence of any heterogeneous genes, we designed immuno-tbid molecules expressing a 10-peptide region (TRHRQPRGWE) derived from amino acids 273 to 282 of PEA to translocate the tBID into the cytoplasm of cultured cells and stimulated the tBID to induce cellular apoptosis. This concept was tested both in vitro and in a mouse model of HER2/neu-positive tumors.

Construction of Expression Plasmids

The RNA samples were extracted from Jurkat cells with TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Reverse transcription-PCR was done to amplify the human bid cDNA. The truncated bid genes, bidΔ1-60 and bidΔ1-75, were obtained by PCR. Using PCR-based overlap extension gene splicing method as described previously (1416), recombinant immuno-tbid was generated by sequential fusion of the genes into a signal peptide (MKHLWFFLLLVAAPRWVLS), a single-chain HER2 antibody (e23sFv, comprising the linked heavy and light chain variable regions derived from a mouse monoclonal antibody (mAb) e23 with high-affinity binding to the extracellular domain of HER2 protein), a PE translocation domain (amino acids 253-358 or 253-364), and a truncated bid gene (bidΔ1-60 or bidΔ1-75). The immuno-tbid gene was cloned into a pCMV plasmid. We also cloned the genes of a partial PEA translocation domain (amino acids 280-358 and 280-364) and fused them to the 5′ end of a tbid gene (tbidΔ1-60 or tbidΔ1-75). The resulting fusion genes were cloned into pcDNA3 and pIRES2-EGFP (BD Biosciences Clontech). Through PCR, immuno-tbid gene was also constructed with a 10-peptide (TRHRQPRGWE) translocation domain, which was derived from amino acids 273 to 282 of PEA, and was cloned into pCMV plasmid. All constructs were verified by DNA sequencing.

Cell Cultures

As reported previously, our human breast cancer (SK-BR-3) and human ovarian cancer (SK-OV-3) are both HER2/neu-overexpressing cells, whereas the human cervical carcinoma (HeLa), African green monkey kidney (COS-7), human umbilical vein endothelial (ECV-304), and human laryngeal cancer (Hep2) are all HER2/neu low-expressing cells (16, 17). The cells were cultured with RPMI 1640 or DMEM (Invitrogen) with 4 mmol/L l-glutamine supplemented with 10% fetal bovine serum and antibiotics. All cells were cultured at 37°C in 95% air-5% CO2.

Immunoblotting

The immuno-tbid transfected cells were cultured with serum-free medium for 3 days and the medium was collected and concentrated for 20 times and then subjected to immunoblotting, which was done following the manufacturer's instructions (Amersham Pharmacia Biotech). Proteins with 15 μg were separated by 10% SDS-PAGE, blotted onto polyvinylidene difluoride membranes, and analyzed with a 1:1,000 dilution of a BID polyclonal goat antibody (C20; Santa Cruz Biotechnologies). Immunoreactive bands were developed with 3,3′-diaminobenzidine staining.

Immunofluorescence

The cells were cultured on coverslips in DMEM containing 10% fetal bovine serum, fixed with a freshly prepared paraformaldehyde solution [4% in PBS (pH 7.4)] for 10 min at room temperature, and permeated with 0.1% Triton X-100 for 15 min. The cells were stained with antibodies recognizing BID (C20, 1:200; Santa Cruz Biotechnologies), AIF (H-300, 1:100; Santa Cruz Biotechnologies), and cytochrome c (H-104, 1:200; Santa Cruz Biotechnologies) as the primary antibodies, with biotin-linked anti-goat IgG (1:100; Santa Cruz Biotechnologies) and FITC-linked anti-rabbit IgG (1:100; Sigma) as the secondary antibodies and Cy3-linked anti-biotin antibodies (1:100; Sigma) as the tertiary antibody. 4′,6-Diamidino-2-phenylindole was also used for nucleus staining.

Immunohistologic Staining

The muscle tissue near the injection spot and tumors were collected and fixed in formalin and were embedded in paraffin after treatment, and the 5 μm sections were cut onto 3-amino-propyltriethoxysilane-coated slides. The slides were dewaxed in xylene and washed in 95% (v/v) ethanol. For immunohistochemistry staining, the muscle tissues were incubated for 30 min in 0.3% hydrogen peroxide solution, washed in PBS, placed in citrate buffer, and microwaved at 900 W for 20 min. All the slides were then dried and blocked for 1 h with the appropriate serum in a humidified chamber. Primary antibody was added overnight at 4°C. BID was analyzed by using a 1:200 dilution of a polyclonal goat antibody, and AIF was analyzed by using a 1:100 dilution of rabbit antibody. The sections were incubated with secondary antibodies for 30 min and then processed with 3,3′-diaminobenzidine staining.

Annexin V-FITC Staining

The transfected HeLa cells were collected, and Annexin V-FITC/PI staining was done according to the manufacturer's instructions (BD PharMingen). Cells were then analyzed by flow cytometry.

APO2.7 Antibody Staining

The cells were washed twice with PBS containing 5% fetal bovine serum, and 1 × 105 cells were suspended in 0.5 mL PBS with 5% fetal bovine serum and 0.02% NaN3 and incubated with mAbs for 30 min at 4°C. For apoptosis detection, we used the mouse phycoerythrin-conjugated mAb APO2.7 (clone 2.7 A6A3; Beckman Coulter) for detecting 7A6 antigen expressed by cells undergoing apoptosis. Flow cytometric analysis was done on a flow cytometer equipped with a 15 mW argon-ion laser with excitation capabilities at 488 nm.

Measurements of Mitochondrial Membrane Potential

The mitochondrial membrane potential of intact cells was measured by flow cytometry with a fluorescent dye that accumulates in mitochondria as a function of membrane potential (3,3′-dihexyloxa-carbocyanine iodide) staining. HeLa cells after transfection were subjected to trypsinization and treated with 3,3′-dihexyloxa-carbocyanine iodide (Invitrogen) in PBS for 15 min at 37°C and subsequently analyzed by fluorescence-activated cell sorting.

Cell-Killing Assay

Cell-killing ratio was determined by cell counting after trypan blue staining, which was used to tell the living cells from the dead ones. The assay was done as follows: (a) detaching the adherent cells with 0.25% (m/v) trypsin (1:250) at 37°C; (b) resuspending the cells in DMEM containing 10% fetal bovine serum in a certain volume; (c) staining the cells with trypan blue to the final concentration of 0.04%; (d) counting respectively the unstained cells in the four squares, each containing 16 smaller squares; (e) calculating the cell concentration to be [(N1 + N2 + N3 + N4) / 4] × 104/mL; and (f) the cell number is that the result of the cell concentration times the total volume, and the cell-killing ratio is that the result of (Nc - Nt) / Nc% in which Nt and Nc are the cell number of treatment and control groups, respectively.

Assessment of In vivo Antitumor Activity of Immuno-tBID

BALB/c athymic mice (6-8 weeks old) were inoculated s.c. with 2 × 106 human breast cancer SK-BR-3 cells. The tumors were allowed to grow and establish until they reached a diameter of 5 to 7 mm (designated day 0). The mice were then divided randomly into groups for the various treatments. The mice bearing SK-BR-3 tumors were subjected to liposome-mediated immuno-tbid gene treatments. The mice were administered i.m. with 10 μg pCMV/immuno-tbid or 10 μg control vector, pCMV, both encapsulated by 20 μL liposomes. Each treatment used 10 mice and was done every 3 days, five times total.

Statistical Analysis

Statistical analyses were done with the SPSS10.0 software package for Windows (SPSS). The survival rates were analyzed using the Kaplan-Meier method, and comparisons among treatment groups were obtained with the log-rank test. Tumor volumes were analyzed with the analysis of covariance, with treatment group comparisons made by covariance testing after eliminating any initial group differences. Statistical significance was based on P < 0.05.

Expression of tBID Contributes to Effective Cell Death

Most of the death agonists in the Bcl2 family, such as BID, contain a BH3 domain that is essential for binding to Bcl2 and capable of inducing apoptosis (1821). Once BID is cleaved, the truncated portion containing the BH3 domain becomes lethal to cells. Two potential cleavage sites reside in the NH2-terminal portion of human BID, 57LQTDG61 and 72IEADS76, which match perfectly with the preferred cleavage sites for caspase-8 and granzyme B, respectively.

To investigate whether the same cytotoxic effect could be achieved in mammalian cells by the expressed truncated BID protein in the cytoplasm, we generated two truncated bid gene (bidΔ1-60 or bidΔ1-75; Fig. 1A) expression plasmids and transiently transfected them into COS-7 and HeLa cells. As shown in Fig. 1B, in contrast to the mock transfectants, the transient expression of the truncated BID (BIDΔ1-60 or BIDΔ1-75) gene led to an apparent delay in cell growth after transfection.

Figure 1.

Cytotoxic effect of BIDΔ1-60/75 and PEA(280-358)-BIDΔ1-60/75 fusion proteins on transfected cells. A,PEA (280-358)-bidΔ1-60/75 fusion genes were constructed by fusing the PEA domain II (amino acids 280-358) in-frame at the 5′ end of the truncated bidΔ1-60/75 sequence. B, growth curves of nontransfected, pCDNA3-transfected, pCDNA3-bidΔ1-60-transfected, and pCDNA3-bidΔ1-75-transfected HeLa cells were assessed by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay with the mean data from four or more independent experiments, each conducted in triplicate. C, HeLa cells transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75 constructs were subjected to terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling staining 24 h later. With microscopy, the apoptotic cells were counted in more than three 100-cell fields and the apoptotic cell ratios were calculated. This experiment was replicated more than four times. D, 12 h after transfection, HeLa cells were subjected to Annexin V-FITC staining concomitant with 4′,6-diamidino-2-phenylindole nucleus staining and analyzed by flow cytometry.

Figure 1.

Cytotoxic effect of BIDΔ1-60/75 and PEA(280-358)-BIDΔ1-60/75 fusion proteins on transfected cells. A,PEA (280-358)-bidΔ1-60/75 fusion genes were constructed by fusing the PEA domain II (amino acids 280-358) in-frame at the 5′ end of the truncated bidΔ1-60/75 sequence. B, growth curves of nontransfected, pCDNA3-transfected, pCDNA3-bidΔ1-60-transfected, and pCDNA3-bidΔ1-75-transfected HeLa cells were assessed by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay with the mean data from four or more independent experiments, each conducted in triplicate. C, HeLa cells transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75 constructs were subjected to terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling staining 24 h later. With microscopy, the apoptotic cells were counted in more than three 100-cell fields and the apoptotic cell ratios were calculated. This experiment was replicated more than four times. D, 12 h after transfection, HeLa cells were subjected to Annexin V-FITC staining concomitant with 4′,6-diamidino-2-phenylindole nucleus staining and analyzed by flow cytometry.

Close modal

PEA (280-358) Fused tBID-Induced Apoptosis by Releasing Cytochrome c and the AIF

PEA is a single-chain toxin consisting of three major domains responsible for binding of the molecule to target cells, its translocation to the cytosol, and the induction of cell death, respectively (11). Indeed the domain II of PEA has been reported to efficiently transfer the cellular toxicity domain to the cytoplasm by undergoing autoprocessing (12). Because endosomal cleavage always occurs between Arg279 and Gly280 residues in the PEA translocation domain, it is necessary to explore whether fusing a portion of the PEA translocation domain (amino acids 280-358) would affect the proapoptotic activity of BIDΔ1-60 and BIDΔ1-75. PEA (280-358)-BIDΔ1-60 or PEA (280-358)-BIDΔ1-75 genes were constructed by adding the encoding sequence of NH2-terminal domain with PEA (280-358) to the BIDΔ1-60 and BIDΔ1-75 genes (Fig. 1A). HeLa cells were transiently transfected with the BIDΔ1-60, BIDΔ1-75, PEA (280-358)-BIDΔ1-60, or PEA (280-358)-BIDΔ1-75 genes. The terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling staining, which is the most commonly used method for detecting DNA degradation in apoptotic cells, was done followed with single blind counting after 24 h (Fig. 1C). These results indicated that all constructs induced cellular apoptosis despite the use of the PEA (280-358) fusion. Moreover, the Annexin V-FITC staining done 12 h after the transfection also revealed that, compared with the control cells, apoptotic cells were present in 11.1%, 10.9%, 11.9%, and 9.2% of the cells transfected with BIDΔ1-60, BIDΔ1-75, PEA (280-358)-BIDΔ1-60, or PEA (280-358)-BIDΔ1-75, respectively (Fig. 1D).

To further show the damage of mitochondria by PEA (280-358)-BIDΔ1-60 and PEA (280-358)-BIDΔ1-75, we studied the expression level of the APO2.7 antigen, which is detectable after mitochondrial pathway apoptosis induction. The APO2.7 antigen (also named 7A6), recognized by clone 2.7A6A3, is a 38-kDa protein localized on the mitochondrial membrane with an expression profile that is restricted to cells undergoing apoptosis. Specifically, APO2.7 expression induction appears as an early event during apoptosis, whereas the normal viable cells are negative or weakly positive for this gene. After transfection with the BIDΔ1-60, BIDΔ1-75, PEA (280-358)-BIDΔ1-60, or PEA (280-358)-BIDΔ1-75 genes, the APO2.7 antigen was presented on the mitochondria surface of HeLa cell (Fig. 2A).

Figure 2.

Cytochrome c and AIF leakage from bidΔ1-60/75- and PEA(280-358)-bidΔ1-60/75-transfected cells. A, HeLa cells were transiently transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75. After 24 h, the cells were subjected to Apo2.7 antibody staining followed by flow cytometry analysis. B, transfected HeLa cells were subjected to trypsinization and analysis of the mitochondrial membrane potential by flow cytometry after 3,3′-dihexyloxa-carbocyanine iodide staining. C, meanwhile, HeLa cells after transfection were subjected to immunofluorescent staining with a mixture of anti-cytochrome c or anti-AIF antibody/FITC and anti-BID antibody/Cy3 together with 4′,6-diamidino-2-phenylindole nucleus staining. Magnification, ×1,000. D, after 30-min exposure to 20 μmol/L zVAD or 20 μmol/L TAT-BH4, HeLa cells were transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75. Then, the cells were subjected to trypan blue staining after 24 h and the negative cells were counted. Accordingly, the cell-killing percentages were calculated.

Figure 2.

Cytochrome c and AIF leakage from bidΔ1-60/75- and PEA(280-358)-bidΔ1-60/75-transfected cells. A, HeLa cells were transiently transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75. After 24 h, the cells were subjected to Apo2.7 antibody staining followed by flow cytometry analysis. B, transfected HeLa cells were subjected to trypsinization and analysis of the mitochondrial membrane potential by flow cytometry after 3,3′-dihexyloxa-carbocyanine iodide staining. C, meanwhile, HeLa cells after transfection were subjected to immunofluorescent staining with a mixture of anti-cytochrome c or anti-AIF antibody/FITC and anti-BID antibody/Cy3 together with 4′,6-diamidino-2-phenylindole nucleus staining. Magnification, ×1,000. D, after 30-min exposure to 20 μmol/L zVAD or 20 μmol/L TAT-BH4, HeLa cells were transfected with pCDNA3, pCDNA3-bidΔ1-60/75, or pCDNA3-PEA(280-358)-bidΔ1-60/75. Then, the cells were subjected to trypan blue staining after 24 h and the negative cells were counted. Accordingly, the cell-killing percentages were calculated.

Close modal

BID is believed to exert its proapoptotic effect by inducing the release of proapoptotic factors, such as cytochrome c and AIF, from the mitochondria. More recently, it has been shown that BID partially permeates the outer mitochondrial membrane and causes the release of several proteins from the intermembrane space. As assayed by staining with an apoptosis-dependent fluorescent dye (3,3′-dihexyloxa-carbocyanine iodide), the mitochondrial membrane potential of the BIDΔ1-60, BIDΔ1-75, PEA (280-358)-BIDΔ1-60, or PEA (280-358)-BIDΔ1-75 gene cells declined, suggesting the possible leakage of internal molecules from mitochondria (Fig. 2B). To further determine cytochrome c and AIF release from the mitochondria, transfected HeLa cells were subjected to immunofluorescent analysis (Fig. 2C). Moreover, the leakage of nuclear materials, which was stained by 4′,6-diamidino-2-phenylindole into the cytoplasm, was observed in the cells expressing BIDΔ1-60, BIDΔ1-75, PEA (280-358)-BIDΔ1-60, or PEA (280-358)-BIDΔ1-75. However, the subsequent apoptosis was inhibited partially by a mitochondria protector TAT-BH4 but not by the caspase pan-inhibitor z-VAD (Fig. 2D), suggesting the activity of AIF-inducing caspase-independent apoptosis. In our study, the expression of truncated BID caused irreversible cell death through apoptosis induction by AIF and cytochrome c release despite the fusion of PEA domain II fragment (amino acids 280-358) at its NH2 terminus.

Selective Killing of Tumor Cells Overexpressing HER2 by the Secreted e23sFv-PEA II (253-358)-BIDΔ1-60/75 Protein

To induce specific tumoricidal effects, we fused the gene encoding an anti-HER2 single-chain antibody (e23sFv) in-frame with the 5′ end of the recombinant genes encoding PEA (253-358) and BIDΔ1-60 or BIDΔ1-75. These new constructions encode the e23sFv-PEA II-BIDΔ1-60/75 molecules with a NH2-terminal signal peptide sequence (ref. 13; Fig. 3A). To investigate the HER2/neu-targeted proapoptotic activity, pCMV/e23sFv-PEA II-BIDΔ1-60/75 constructs were transfected into HeLa cells, and the cell clones that stably expressed the target fusion proteins were generated. The existence of the secreted e23sFv-PEA II-BIDΔ1-60/75 proteins was confirmed in the transiently transfected HeLa and SK-BR-3 cells by immunofluorescence staining (Fig. 3B). As shown in Fig. 3B, positive signal presented not only around the nucleus of HeLa and SK-BR-3 cells, suggesting immuno-tBID would be secretively expressed, but also in the cytoplasm of HER2/neu-positive SK-BR-3 cells, implying its reentry through cell membrane. Moreover, the culture medium of the e23sFv-PEA II-BIDΔ1-60/75 modified HeLa cells were collected and the presence of e23sFv-PEA II-BIDΔ1-60/75 proteins was determined (Fig. 3C).

Figure 3.

Immuno-tBID expressed secretively. A,immuno-tbid constructs were generated by fusing the e23sFv gene in-frame with the 5′ end of the fusion genes encoding the PEA domain II (amino acids 280-358) and bidΔ1-60/75. B, HeLa and SK-BR-3 cells were transiently transfected with e23sFv-PEA II-bidΔ1-60, respectively, and immunofluorescent staining was done with anti-BID antibody/Cy3. Magnification, ×1,000. C, HeLa cells were transfected with e23sFv-PEA II-bidΔ1-60 and e23sFv-PEA II-bidΔ1-75 followed by 800 μg/mL G418 selecting. The medium of 1 × 106e23sFv-PEA II-bidΔ1-60/75 gene-modified HeLa cells after 48 h of cultivation was concentrated and subjected to immunoblotting with a BID antibody. Lane 1, protein marker; lane 2, vector; lane 3, e23sFv-PEA II-bidΔ1-60; lane 4, e23sFv-PEA II-bidΔ1-75.

Figure 3.

Immuno-tBID expressed secretively. A,immuno-tbid constructs were generated by fusing the e23sFv gene in-frame with the 5′ end of the fusion genes encoding the PEA domain II (amino acids 280-358) and bidΔ1-60/75. B, HeLa and SK-BR-3 cells were transiently transfected with e23sFv-PEA II-bidΔ1-60, respectively, and immunofluorescent staining was done with anti-BID antibody/Cy3. Magnification, ×1,000. C, HeLa cells were transfected with e23sFv-PEA II-bidΔ1-60 and e23sFv-PEA II-bidΔ1-75 followed by 800 μg/mL G418 selecting. The medium of 1 × 106e23sFv-PEA II-bidΔ1-60/75 gene-modified HeLa cells after 48 h of cultivation was concentrated and subjected to immunoblotting with a BID antibody. Lane 1, protein marker; lane 2, vector; lane 3, e23sFv-PEA II-bidΔ1-60; lane 4, e23sFv-PEA II-bidΔ1-75.

Close modal

We next cultured HER2/neu-overexpressing SK-BR-3 and SK-OV-3 cells or HER2/neu low-expressing Hep-2 and ECV-304 cells in the medium of HeLa/e23sFv-PEA II-BIDΔ1-60/75 cells to evaluate whether e23sFv-PEA II-BIDΔ1-60/75 exerted a potent in vitro specific cytotoxicity on tumor cells. We also examined which of the molecules wielded the more powerful death-inducing activity and when the killing rate peaked. The cell-counting process showed that the HeLa/e23sFv-PEA II-BIDΔ1-60/75 cell medium killed the HER2/neu-positive tumor cells (SK-BR-3 and SK-OV-3) specifically but not the Hep-2 or the ECV-304 cells (Fig. 4B). The cell-killing activity was enhanced by prolonged incubation periods (Fig. 4B), inhibited partially by zVAD, and nearly blocked by TAT-BH4 (Fig. 4C and D). In addition, between the two types, e23sFv-PEA II-BIDΔ1-60, with a killing ratio of ∼81.5% after 5 days of treatment, was the more efficient killer protein (Fig. 4A), leading us to choose this molecule for additional studies. These results indicated that secreted e23sFv-PEA II-BIDΔ1-60/75 can recognize and destroy HER2-overexpressing tumor cells, but none of the cell lines lacked the ability to express HER2 (including HeLa cells).

Figure 4.

Selective cytotoxicity of immuno-tBID-secreting Jurkat cells toward tumor cells. A, HER2-positive tumor cells (SK-BR-3) were cultured with the conditioned medium of e23sFv-PEA II-bidΔ1-60/75-modified HeLa cells and counted. The cell-killing percentages were calculated over 5 d from the cell counts. B, SK-BR-3, SK-OV-3, Hep-2, and ECV-304 were counted after 1 to 6 d of culture in the medium of HeLa cells transfected with e23sFv-PEA II-bidΔ1-60, and the percentages of killed cells were determined. With or without the addition of 20 μmol/L zVAD and 20 μmol/L TAT-BH4, the SK-BR-3 cells were cultivated in the conditioned medium of e23sFv-PEA II-bidΔ1-60 (C) and immuno-caspase-3, immuno-caspase-6, and immuno-granzyme B modified cells (D), respectively, and the killing percentages were calculated.

Figure 4.

Selective cytotoxicity of immuno-tBID-secreting Jurkat cells toward tumor cells. A, HER2-positive tumor cells (SK-BR-3) were cultured with the conditioned medium of e23sFv-PEA II-bidΔ1-60/75-modified HeLa cells and counted. The cell-killing percentages were calculated over 5 d from the cell counts. B, SK-BR-3, SK-OV-3, Hep-2, and ECV-304 were counted after 1 to 6 d of culture in the medium of HeLa cells transfected with e23sFv-PEA II-bidΔ1-60, and the percentages of killed cells were determined. With or without the addition of 20 μmol/L zVAD and 20 μmol/L TAT-BH4, the SK-BR-3 cells were cultivated in the conditioned medium of e23sFv-PEA II-bidΔ1-60 (C) and immuno-caspase-3, immuno-caspase-6, and immuno-granzyme B modified cells (D), respectively, and the killing percentages were calculated.

Close modal

HER2/neu-Positive Tumor Growth Was Suppressed by e23sFv-PEA II-bidΔ1-60 Transduction In vivo

The 30 mice with xenografts were divided randomly into three groups to receive i.m. injections of liposome-encapsulated pCMV/e23sFv-PEA II-BIDΔ1-60, pCMV vector alone, or PBS-liposome mixture, respectively. The tumor growth and survival rates were traced and analyzed. The tumors treated with the liposome-pCMV/e23sFv-PEA II-BIDΔ1-60 complex grew more slowly than those treated with the liposome-pCMV complex (Fig. 5B), which suggested that the e23sFv-PEA II-BIDΔ1-60 protein could strongly suppress the HER2/neu-positive tumor on secretion by genetically modifying the tumor cells themselves. Meanwhile, the mice treated with the pCMV/e23sFv-PEA II-BIDΔ1-60 construct exhibited prolonged survival time compared with those treated with a control vector (Fig. 5B; P < 0.05).

Figure 5.

Suppression of HER2-overexpressing tumor cells in vivo by immuno-tBID. The SK-BR-3-recipient (2 × 106) nude mice received i.m. injections of 10 μg LipofectAMINE-encapsulated pCMV-e23sFv-PEA II-bidΔ1-60 or pCMV plasmid in the right posterior limb every 3 d for a total of five times and then once a week (10 mice per group) thereafter. The tumor volume was monitored, and the injected muscle tissues were collected and subjected to immunohistochemistry staining with the anti-BID antibody (A). From these results, the mouse survival times were determined (B). The tumor tissues were collected and subjected to immunofluorescent staining with a mixture of anti-AIF antibody/FITC and anti-BID antibody/Cy3 together with 4′,6-diamidino-2-phenylindole nucleus staining (C) and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling staining (D).

Figure 5.

Suppression of HER2-overexpressing tumor cells in vivo by immuno-tBID. The SK-BR-3-recipient (2 × 106) nude mice received i.m. injections of 10 μg LipofectAMINE-encapsulated pCMV-e23sFv-PEA II-bidΔ1-60 or pCMV plasmid in the right posterior limb every 3 d for a total of five times and then once a week (10 mice per group) thereafter. The tumor volume was monitored, and the injected muscle tissues were collected and subjected to immunohistochemistry staining with the anti-BID antibody (A). From these results, the mouse survival times were determined (B). The tumor tissues were collected and subjected to immunofluorescent staining with a mixture of anti-AIF antibody/FITC and anti-BID antibody/Cy3 together with 4′,6-diamidino-2-phenylindole nucleus staining (C) and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling staining (D).

Close modal

Immunohistochemical or immunofluorescence analysis confirmed BID-positive staining in muscles (Fig. 5A) and tumors (Fig. 5C) treated with pCMV-e23sFv-PEA II-BIDΔ1-60 but not in those treated with pCMV vector alone. Apoptotic cells were discovered using the terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling assay in tumor tissues treated with liposome-pCMV/e23sFv-PEA II-BIDΔ1-60 (Fig. 5D).

Selective Cell-Killing Effect Induced by e23sFv-TP-BIDΔ1-60 In vitro and In vivo

Our study shows that the chimeric protein including the PEA II (253-358) translocation domain has a potent selective antitumor effect both in vitro and in vivo. This result prompted us to shorten the PEA translocation domain and transfect the PEA-amended constructs into appropriate cell lines. We engineered these e23sFv-TP-bidΔ1-60 constructs by fusing the e23sFv gene in-frame with the 5′ end of the fusion genes encoding the PEA domain II (amino acids 273-282) and BIDΔ1-60. We evaluated its antitumor activity in vitro and then constructed and analyzed similarly a recombinant gene without a translocation domain as a negative control. This was constructed by fusing the antibody gene against HER2 with the 5′ end of the BIDΔ1-60 gene directly (Fig. 6A).

Figure 6.

Ten peptides of PEA translocation domain did not affect the apoptotic induction activity of immuno-tBID. A,e23sFv-TP-bidΔ1-60 construct was generated by fusing the e23sFv gene in-frame with the 5′ end of the fusion genes encoding the PEA domain II (amino acids 273-282) and BIDΔ1-60. B, SK-BR-3 cells were transiently transfected with pCMV-e23sFv-TP-bidΔ1-60, e23sFv-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, or the pCMV vector and harvested for 75% ethanol fixing 2 d after, and the cell cycle was analyzed by flow cytometry. C, additionally, the Jurkat cells were transfected with pCMV-e23sFv-TP-bidΔ1-60, e23sFv-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, or the pCMV vector followed by G418 selection (800 μg/mL) and were cocultivated with SK-BR-3 cells on 12-well Costar Transwell at a ratio of 1:1 for 1 to 5 d, and the killing percentages were calculated with the proof of cell counting after trypan blue staining. D, SGC7901-recipient (1 × 106) nude mice received i.m. injections of 10 μg LipofectAMINE-encapsulated pCMV-e23sFv-TP-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, caspase-3, caspase-6, and granzyme B constructs or the pCMV plasmid in the right posterior limb every 3 d for five times (10 mice per group) and then once a week thereafter. The mouse survival times were traced and recorded.

Figure 6.

Ten peptides of PEA translocation domain did not affect the apoptotic induction activity of immuno-tBID. A,e23sFv-TP-bidΔ1-60 construct was generated by fusing the e23sFv gene in-frame with the 5′ end of the fusion genes encoding the PEA domain II (amino acids 273-282) and BIDΔ1-60. B, SK-BR-3 cells were transiently transfected with pCMV-e23sFv-TP-bidΔ1-60, e23sFv-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, or the pCMV vector and harvested for 75% ethanol fixing 2 d after, and the cell cycle was analyzed by flow cytometry. C, additionally, the Jurkat cells were transfected with pCMV-e23sFv-TP-bidΔ1-60, e23sFv-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, or the pCMV vector followed by G418 selection (800 μg/mL) and were cocultivated with SK-BR-3 cells on 12-well Costar Transwell at a ratio of 1:1 for 1 to 5 d, and the killing percentages were calculated with the proof of cell counting after trypan blue staining. D, SGC7901-recipient (1 × 106) nude mice received i.m. injections of 10 μg LipofectAMINE-encapsulated pCMV-e23sFv-TP-bidΔ1-60, e23sFv-PEA II-bidΔ1-60, caspase-3, caspase-6, and granzyme B constructs or the pCMV plasmid in the right posterior limb every 3 d for five times (10 mice per group) and then once a week thereafter. The mouse survival times were traced and recorded.

Close modal

The pCMV/e23sFv-PEA II-BIDΔ1-60, e23sFv-TP-BIDΔ1-60, and e23sFv-BIDΔ1-60 were stably transfected into Jurkat and CHO cells, respectively, followed by G418 selection. The G418-resistant cells were subcloned, and the immunoblotting of the culture medium confirmed the stable secretion of the chimeric proteins. We showed through fluorescence microscopy that after transiently transfecting these genes into the HER2-positive breast cancer SK-BR-3 and gastric cancer SGC7901 cells under LipofectAMINE mediation, sub-G1 peaks were appeared nearly 48 h later (Fig. 6B) in e23sFv-PEA II-BIDΔ1-60 and e23sFv-TP-BIDΔ1-60 transfectant groups. Moreover, after being cocultivated with e23sFv-PEA II-BIDΔ1-60 or e23sFv-TP-BIDΔ1-60 modified Jurkat cells, the SK-BR-3 cells died in an apoptotic manner (Fig. 6C).

To investigate the efficiency of the e23sFv-TP-BIDΔ1-60 fragment and the diversity of several immuno-proapoptotic proteins, the SGC7901 recipient (1 × 106) nude mice were injected i.m. with 10 μg LipofectAMINE-encapsulated pCMV/e23sFv-TP-BIDΔ1-60, e23sFv-PEA II-BIDΔ1-60, immuno-caspase-3, immuno-caspase-6, or immuno-granzyme B, and pCMV as a control. The result showed that e23sFv-TP-BIDΔ1-60 and e23sFv-PEA II-BIDΔ1-60 prolonged the mouse survival time remarkably. Moreover, this effect was higher than that observed for immuno-granzyme B and similar to that observed for immuno-caspase-3 and immuno-caspase-6 (Fig. 6D).

The HER2 oncogene, also called ErbB-2/neu, is the human homologue of the neu oncogene identified in DNA from rat neurogloblastomas as induced by ethylnitrosourea (22). Located on chromosome 17q, HER2 encodes a transmembrane glycoprotein (p185) that was classified as a member of the epidermal growth factor receptor family. The HER2 protein is reportedly overexpressed in several adenocarcinomas, including breast, ovarian, lung, and stomach (26). Sufficient evidence has suggested that patients with HER2-overexpressing tumors exhibit a reduced response to conventional treatments (23). Because it is overexpressed in tumor cells but is not detected in normal cells, HER2 is an ideal target molecular for cancer gene therapy to exploit differences at the molecular level between normal and malignant cells.

Several new therapeutic approaches have been developed that are based on mAbs or on antisense technology targeting the cells overexpressing HER2. Humanized anti-HER2 mAbs alone or conjugated either to toxins or enzymes directed against the extracellular domain of HER2 have been shown to be active in metastatic cancers overexpressing HER2 in xenograft mice (10, 2426). Many early studies noted the improvement in limiting tumor growth in HER2-positive cell lines using an immunotoxin constructed by single-chain Fv modified PEA to convert from a cytostatic to a cytotoxic effect both in vitro and in vivo. Unfortunately, the in vivo use of immunotoxins for cancer therapy is not entirely advantageous in its current form. Reportedly, administration of analogous immunotoxins can result in the development of allergic reactions, infusion reactions, and lung problems in clinical experiments (26, 27). This is likely due to the heterogeneous nature of these toxins.

The most promising strategy for treating HER2-positive tumors will likely combine mAbs targeting of cancer cells with a maximal cell-killing ability while minimizing the potential damage to the surrounding normal tissues. Our approach to this end is using the anti-HER2 single-chain Fv targeted to tumor cells and with the minimized translocation domain to introduce a human gene encoding a key proapoptotic protein into the cytoplasm. Previously, several proapoptotic molecules, including caspase-3, caspase-6, granzyme B, and AIF were constructed into our immuno-proapoptotic proteins and were determined to be efficient in vitro and in vivo. Herein, we introduced tBID into this strategy by engineering several immuno-tbid constructs.

BID is a proapoptotic, BH3 domain–only member of the Bcl-2 family. As a key player in the apoptosis pathway, BID is a cytosolic protein in nonapoptotic cells, presenting in a variety of tissues (28). On induction of certain types of apoptosis, BID is cleaved by caspase-8, and its COOH-terminal fragment translocates to the mitochondria, where it exerts its proapoptotic effect by inducing the release of proapoptotic factors, such as cytochrome c, AIF, and procaspase 9 (8, 9). The COOH-terminal fragment of BID (called tBID) is regarded as the active form, which is far more potent in inducing apoptosis than the full-length BID molecule.

The tBID molecule was modified into a novel chimeric protein named immuno-tbid. This construction was achieved by fusing tBID to an anti-HER2 single-chain antibody and a NH2-terminal PEA translocation domain. In theory, the e23sFv-PEA II-BIDΔ1-60/75 molecule can bind to and enter HER2-overexpressing tumor cells. To select the tBID construct with the greater tumor-killing potential, we evaluated two versions of active BID (BIDΔ1-60 and BIDΔ1-75).

Given that endocytosis of the e23sFv-PEA II-BIDΔ1-60/75 protein by HER2-overexpressing tumor cells would result in the release of PEA II-tBID to the cytosol, we investigated whether the PEA-derived peptide affected the tBID activity. Based on the well-characterized translocation mechanism of the PEA domain II in chimeric toxins (11, 12), autocleavage of the PEA translocation domain between Arg279 and Gly280 residues in the target cell endosomes would give rise to a COOH-terminal fragment, termed PEA (280-358)-BIDΔ1-60/75. The resulting fragment would consequently translocate to the cytosol and induce the cell to undergo apoptosis. The data from both transient and inducible expression systems showed that the PEA II-tBID form was active and caused cell death effectively in a manner similar to that seen with the tbid transfectants.

Furthermore, by cultivating in the medium of the genetically modified HeLa cells, our in vitro and in vivo data showed significant killing of the HER2-overexpressing tumor cells but not the HER2 low-expressing tumor cells or normal cells. In contrast to the immuno-apoptotic proteins that we generated previously (1317), the e23sFv-PEA II-BIDΔ1-60/75 molecule exerts only minor immunogenicity potential and therefore should be applicable to long-term treatment of tumors that overexpress HER2. Moreover, unlike our immuno-caspase-3, immuno-caspase-6, or immuno-granzyme B constructs, e23sFv-PEA II-BIDΔ1-60/75 would be useful generally when either the receptor or mitochondrial-mediated apoptotic pathway is not available in tumor cells.

Because of the uncertainty of whether the minimal fragment of PEA is necessary for its translocation activity, we constructed a short translocation peptide to link e23sFv and tBID with only 10 amino acids derived from the furin cleavage site (29) of PEA, called e23sFv-TP-BIDΔ1-60, and compared its antitumor effects with those of immuno-apoptotic proteins in vitro and in vitro. Importantly, the molecules inhibited cell proliferation similarly and killed tumor cells overexpressing HER2. We also compared the in vivo effects of the immuno-tBID and immuno-TP-tBID constructs with those of other immuno-proapoptotic proteins that we constructed previously. The result showed that immuno-tBID and immuno-TP-tBID could inhibit tumor growth more strongly and prolong longevity in mice. This advantage may reflect a smaller influence of the shorter NH2-terminal PEA-derived peptide on tBID activity and on the ability of the molecule to escape from endosomes.

A major concern in the development of HER2-targeted therapies is the expression of the target antigen on normal tissues. Numerous studies with Herceptin (trastuzumab), although showing significant clinical benefits in patients with HER2-positive primary and metastatic breast cancers, have reported notable cardiotoxicity in a few patients presumably because of interference with normal HER2 signaling in the heart (30). By contrast, immuno-TP-tBID fusion proteins are expected to be less toxic over repeated treatments because of enhanced antitumor activity through the addition of an apoptotic effect molecule that permits decreased doses of the targeted killer protein to be used in vivo.

No potential conflicts of interest were disclosed.

Grant support: National Basic Research Program of China grant 2004CB518805, National High Technology Research and Development Program of China grants 2006AA02A255 and 2007AA021104, National Science Foundation of China grant 30471988, Program for Changjiang Scholars and Innovative Research Team in University grant IRT0459, and China Postdoctoral Science Foundation grant 2005038259.

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.

Note: X-C. Qiu and Y-M. Xu contributed equally to this article.

We thank Drs. Dai-ming Fan and Yong Zhou for useful suggestions and encouragement and Yun-xin Cao for assistance with the flow cytometric analysis.

1
Lupu R, Colomer R, Kannan B, et al. Characterization of a growth factor that binds exclusively to the erbB-2 receptor and induces cellular responses.
Proc Natl Acad Sci U S A
1992
;
89
:
2287
–91.
2
Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer.
Science
1989
;
244
:
707
–12.
3
Semba K, Kamata N, Toyoshima K, et al. A v-erbB related protooncogene, c-erbB-2, is distinct from the c-erbB-1/epidermal growth factor-receptor gene and is amplified in a human salivary gland adenocarcinoma.
Proc Natl Acad Sci U S A
1985
;
82
:
6497
–501.
4
Fukushige S, Matsubara K, Yoshida M, et al. Localization of a novel v-erbB-related gene, c-erbB-2, on human chromosome 17 and its amplification in a gastric cancer cell line.
Mol Cell Biol
1986
;
6
:
955
–8.
5
Eccles SA. The role of c-erbB-2/HER2/neu in breast cancer progression and metastasis.
J Mammary Gland Biol Neoplasia
2001
;
6
:
393
–406.
6
Yarden Y. Biology of HER2 and its importance in breast cancer.
Oncology
2001
;
61
:
1
–13.
7
Yin XM. Signal transduction mediated by Bid, a pro-death Bcl-2 family protein, connects the death receptor and mitochondria apoptosis pathways.
Cell Res
2000
;
10
:
161
–7.
8
Li H, Zhu H, Xu CJ, et al. Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis.
Cell
1998
;
94
:
491
–501.
9
Luo X, Budlhardjo I, Zou H, et al. Bid, a Bcl-2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors.
Cell
1998
;
94
:
481
–90.
10
Batra JK, Kasprzyk PG, Bird RE, et al. Recombinant anti-erbB2 immunotoxins containing Pseudomonas exotoxin.
Proc Natl Acad Sci U S A
1992
;
89
:
5867
–71.
11
Hwang J, Fitzgerald DJ, Adhya S, et al. Functional domains of Pseudomonas exotoxin identified by deletion analysis of the gene expressed in E. coli.
Cell
1987
;
48
:
129
–36.
12
Siegall CB, Chaudhary VK, FitzGerald DJ, et al. Functional analysis of domains II, Ib, and III of Pseudomonas exotoxin.
J Biol Chem
1989
;
264
:
14256
–61.
13
Yang AG, Chen SY. A new class of antigen-specific killer cells.
Nat Biotechnol
1997
;
15
:
46
–51.
14
Jia LT, Zhang LH, Yu CJ, et al. Specific tumoricidal activity of a secreted proapoptotic protein consisting of HER2 antibody and constitutively active caspase-3.
Cancer Res
2003
;
63
:
3257
–62.
15
Xu YM, Wang LF, Jia LT, et al. A caspase-6 and anti-HER2 antibody chimeric molecule suppresses the growth of HER2-overexpressing tumor.
J Immunol
2004
;
173
:
61
–7.
16
Zhao J, Zhang LH, Jia LT, et al. Secreted antibody/granzyme B fusion protein stimulates selective killing of HER2-overexpressing tumor cells.
J Biol Chem
2004
;
279
:
21343
–8.
17
Yu CJ, Jia LT, Meng YL, et al. Selective proapoptotic activity of a secreted recombinant antibody/AIF fusion protein in carcinomas overexpressing HER2.
Gene Ther
2006
;
13
:
313
–20.
18
Chittenden T, Flemington C, et al. A conserved domain in Bak, distinct from BH1 and BH2, mediates cell death and protein binding functions.
EMBO J
1995
;
14
:
5589
–96.
19
Wang K, Yin XM, Chao DT, et al. BID: a novel BH3 domain-only death agonist.
Genes Dev
1996
;
10
:
2859
–69.
20
Kelekar A, Chang BS, Harlan JE, et al. Bad is a BH3 domain-containing protein that forms an inactivating dimer with Bcl-XL.
Mol Cell Biol
1997
;
17
:
7040
–6.
21
Cosulich SC, Worrall V, Hedge PJ, et al. Regulation of apoptosis by BH3 domains in a cell-free system.
Curr Biol
1997
;
7
:
913
–20.
22
Schechter AL, Stern DF, Vaidyanathan L, et al. The neu oncogene: an erb-B-related gene encoding a 185,000-Mr tumour antigen.
Nature
1984
;
312
:
513
–6.
23
Chen JS, Lan K, Hung MC. Strategies to target HER2/neu overexpression for cancer therapy.
Drug Resist Updat
2003
;
6
:
129
–36.
24
Wels W, Harwerth IM, Mueller M, et al. Selective inhibition of tumor cell growth by a recombinant single-chain antibody specific for the erbB-2 receptor.
Cancer Res
1992
;
52
:
6310
–7.
25
Schmidt M, Hynes NE, Groner B, et al. A bivalent single chain antibody-toxin specific for ErbB-2 and the EGF receptor.
Int J Cancer
1996
;
65
:
538
–46.
26
Azemar M, Djahansouzi S, Jager E, et al. Regression of cutaneous tumor lesions in patients intratumorally injected with a recombinant single-chain antibody-toxin targeted to ErbB2/HER2.
Breast Cancer Res Treat
2003
;
82
:
155
–64.
27
Pai-Scherf LH, Villa J, Pearson D, et al. Hepatotoxicity in cancer patients receiving erb-38, a recombinant immunotoxin that targets the erbB2 receptor.
Clin Cancer Res
1999
;
5
:
2311
–5.
28
Yin XM. Bid, a critical mediator for apoptosis induced by the activation of Fas/TNF-R1 death receptors in hepatocytes.
J Mol Med
2000
;
78
:
203
–11.
29
Thomas G. Furin at the cutting edge: from protein traffic to embryogenesis and disease.
Nat Rev Mol Cell Biol
2002
;
3
:
753
–66.
30
Le XF, Bedrosian I, Mao W, et al. Anti-HER2 antibody trastuzumab inhibits CDK2-mediated NPAT and histone H4 expression via the PI3K pathway.
Cell Cycle
2006
;
5
:
1654
–61.