Herceptin Sensitizes ErbB2–Overexpressing Cells to Apoptosis by Reducing Antiapoptotic Mcl-1 Expression

The epidermal growth factor (EGF) receptor ErbB2 (also known as HER2/neu) is a 185 kDa transmembrane receptor tyrosine kinase that is amplified or overexpressed in many cancers including breast cancer (1). It has been associated with poor clinical outcome such as shorter survival and short time to relapse (2, 3). This has led to ErbB2 being a target for drug development. Monoclonal antibodies against ErbB2 (herceptin, Trastuzumab) have been developed which reduces tumor growth and induces apoptosis (1, 4, 5). Herceptin is currently used in the treatment of breast tumors expressing high levels of ErbB2 (5). This treatment is most effective when combined with standard chemotherapy such as microtubule toxins (taxol) or genotoxins (etoposide; ref. 6). The mechanism for the synergy between herceptin and standard chemotherapy remains unclear.

Bcl-2 family members consist of both proapoptotic (Bax and Bak) and antiapoptotic (Bcl-2, Bcl-x_L, A1, Mcl-1, and Bcl-w) proteins (7). The association of these proteins with each other dictates whether a cell will survive or undergo apoptosis (8). In cancer, antiapoptotic members are often increased, rendering cancer cells resistant to apoptosis (8). Mcl-1 is one member of the antiapoptotic Bcl-2 family that was initially identified in the human myeloid leukemia cell line, ML-1 (9, 10). The protein inhibits apoptosis, through a direct interaction with proapoptotic members of the Bcl-2 family at the mitochondria. Mcl-1 is essential for hematopoietic stem cells and lymphocyte survival (11). The Mcl-1 protein has a hydrophobic stretch of 20 amino acids at the COOH terminus, which is responsible for localizing both proteins to the outer mitochondrial membrane similar to other Bcl-2 family members. Mcl-1 differs from Bcl-2 at the amino end where Mcl-1 has a unique amino acid segment that contains PEST motifs (sequences rich in proline, glutamate, serine, and threonine; refs. 9, 10). These PEST motifs likely contribute to the fact that the half-life of Mcl-1 is ∼1 hour, compared with 10 to 14 hours for Bcl-2. Mcl-1 mRNA is also alternatively spliced into a short form (Mcl-1s; ref. 12). In addition, Mcl-1 mRNA turns over rapidly, and these features explain the ability of Mcl-1 expression to be rapidly increased or down-regulated with cytokines and differentiation factors (13). Mcl-1 thus differs from other members of the Bcl-2 family in acting as an immediate response molecule to protect cells against apoptosis.

Herein, we will show that ErbB2 expressing cells up-regulate Mcl-1 levels and upon herceptin treatment, Mcl-1 expression is reduced. This corresponds with increased sensitivity in these cells to apoptosis. Using antisense against Mcl-1, MDA-MB-231 breast cancer cells are sensitized to etoposide-induced apoptosis that is not enhanced by further herceptin treatment. This suggests that herceptin antitumor activity could be due to reduced expression of Mcl-1.

Materials. Antibodies for Western blot were obtained from Santa Cruz Biotechnology (Santa Cruz, CA): Mcl-1 rabbit polyclonal S-19 (sc-819), Bcl-x_S/L rabbit polyclonal S-18 (sc-634), and Bcl-2 rabbit polyclonal N-19 (sc-492). Actin rabbit polyclonal (A02066), anti-rabbit FITC (F-0382) and Bax mouse monoclonal clone 2D2 (B-8554) were purchased from Sigma-Aldrich (St. Louis, MO). Antibodies for immunofluorescence were obtained from DAKO Corporation (Carpinteria, CA): Mcl-1 rabbit polyclonal (A3534), c-erbB-2 rabbit polyclonal (A0485), swine anti-rabbit blocking antibody (Z0196). Anti-mouse Rhodamine Red-X (R-6393) was purchased from Molecular Probes (Eugene, OR).

Cell culture. All cell lines (unless otherwise stated) were maintained in a humidified 5% CO₂ environment, in DMEM supplemented with 100 units of penicillin per mL, 100 μg of streptomycin (Life Technologies) per mL and in 10% bovine calf serum. Cells were transfected with Mcl-1 cDNA by LipofectAMINE as previously described (14). NIH3T3 cells contain very low levels of expression of EGF receptors and either untransfected or stably transfected with human ErbB2 cDNA in an expression vector as previously described (15). MDA-MB-231 cells were derived from a metastasis in mice (a kind gift from Dr. Peter Watson, Manitoba Institute of Cell Biology and Department of Pathology, University of Manitoba, Winnipeg, Manitoba, Canada). MCF-7 cells were maintained in α-MEM with 10% fetal bovine serum (Invitrogen, Carlsbad, CA), and 1% penicillin/streptomycin (Invitrogen), HEPES (Invitrogen), and sodium pyruvate (Invitrogen). Cells were visualized using an Olympus CK40 microscope and Spot CCD Camera (Diagnostic Instruments, Inc., Sterling Heights, MI).

Measurement of apoptosis. Cells were resuspended in 100 μL of medium and 2 μL of acridine orange (100 μg/mL), and ethidium bromide (100 μg/mL) in PBS was added. Ten microliters of the solution were applied to a slide, and the cells were viewed on an Olympus BX51 fluorescent microscope using the fluorescein filter set. Cells which showed intense staining of the DNA in the nucleus were defined as apoptotic versus the diffuse staining of the nucleus in healthy cells. At least 200 cells were counted per condition. All statistical analyses for significant differences in apoptotic responses was done using a Student's t test method unless otherwise stated.

RNase protection assay. A RiboQuant Multi-Probe RNase Protection Assay system (BD Biosciences PharMingen, Mississauga, Ontario, Canada) was used as per the manufacturer's instructions. MDA-MB-231 cells were treated for 24 hours with 1 or 10 μg of Herceptin, or left untreated as a control. RNA was harvested using RNA Bee (Tel-Test Inc., Friendswood, TX) as per the manufacturer's instructions and 20 μg of RNA was hybridized with the hAPO-2b template set (BD Biosciences PharMingen). The housekeeping gene GAPDH was used to normalize the samples. Analysis of the signal intensity was done on a Storm Phosphoimager (GE Health Care Biosciences, Little Chalfont, United Kingdom).

Statistical analysis. Statistical analysis was done using Microsoft Excel, and using Student's t test or χ² tests as appropriate. P < 0.05 was considered statistically significant (16).

Immunoblots. Cells were lysed in Nonidet P-40 lysis buffer [50 mmol/L HEPES (pH 7.25), 150 mmol/L NaCl, 50 μmol/L ZnCl₂, 50 μmol/L NaF, 2 mmol/L EDTA, 1.0% Nonidet P-40, 2 mmol/L phenylmethylsulfonyl fluoride]. Cell debris was removed by centrifugation at 10,000 × g for 10 minutes and protein concentrations were determined by Bradford assay. Fifty micrograms of lysate were subjected to SDS-polyacrylamide electrophoresis and transferred to nitrocellulose membranes. The membranes were blocked in TBS with 0.15% Tween 20 (TBST) and 5% milk. Blots were incubated with a primary antibody overnight, washed thrice in 1× TBST and incubated for 1 hour with the appropriate secondary antibody conjugated with alkaline phosphatase. Blots were visualized on X-ray film with enhanced chemiluminescence reagents (Amersham-Biosciences). Breast tumor protein was a kind gift from the Manitoba Breast Tumor Bank. Equal amounts of protein were loaded in each lane as confirmed by Ponceau-S staining.

Transfection of antisense oligonucleotides against Mcl-1. All oligonucleotides (Sigma Genosys) were designed with phosphorothioate nucleotides at each end represented by the lower case letters in the sequence. Antisense (catccCAGCCTCTTtgtttA), sense (atttgTTTCTCCGAccctaC) and scramble (tacgtTCGTTTCCAcctcaT) oligonucleotides were used. The scrambled sequence was run through BLAST search, and no significant homologies existed with other genes. Antisense transfections were done using OligofectAMINE Reagent (Invitrogen) as previously described (14). At this time, 10 μg/mL of Herceptin was added to the appropriate wells and the cells were incubated for 1 hour before etoposide was added. Cells were harvested 24 hours after etoposide treatment.

Immunofluorescence on breast tumor sections. Paraffin-embedded sections were provided by the Manitoba Breast Tumor Bank. Immunofluorescence was done as described in ref. (17). Briefly, anti-Mcl-1 (DAKO) was diluted 1:250 in blocking buffer [1× PBS, 5% goat serum (Life Technologies), 0.2% Triton X-100, 0.02% sodium azide, and 0.1% bovine serum albumin]. All incubations in primary antibody were done overnight, incubations in secondary antibodies were done in 1 hour. Secondary anti-rabbit Rhodamine Red was diluted at 1:500 in blocking buffer. Because both primary antibodies were raised in rabbit, additional blocking was required. Slides were blocked in 100% bovine calf serum for 30 minutes, washed thrice for 5 minutes, and blocked with swine anti-rabbit (Sigma) at a dilution of 1:25 in blocking buffer for 30 minutes. Anti-ErbB2 was diluted 1:350 in blocking buffer. Secondary anti-rabbit FITC was diluted 1:500 in blocking buffer. Nuclei were stained using Hoechst, and the slides were mounted using antifade (Molecular Probes). Fluorescence was visualized and captured on an Olympus BX51 microscope (Olympus) with Scope Pro deconvolution software (Media Cybernetics).

Breast cancer cell lines are sensitized to etoposide- and taxol-induced apoptosis following herceptin treatment. Increased ErbB2 expression has been associated with drug resistance in breast cancer (1, 2). In breast cancer patients with overexpression of ErbB2, herceptin treatment is often used in combination with standard chemotherapy (4, 5). To determine whether herceptin could sensitize breast cancer cells to apoptosis, MDA-MB-231 cells were treated with etoposide (1.5 mmol/L) in the presence or absence of herceptin (0.1 to 10 μg/mL) for 48 hours. MDA-MB-231 cells are relatively resistant to etoposide-induced apoptosis in which 1.5 mmol/L etoposide treatment induced only 26% apoptosis (Fig. 1A). When 0.5 μg/mL herceptin was added, the amount of etoposide-induced apoptosis increased to 29% and when 1 or 10 μg/mL herceptin was added, the amount of apoptosis increased to 36% and 39%, respectively (Fig. 1A). Similar results were found in the breast cancer cell line MCF-7 when 10 μg/mL herceptin significantly increased the amount of etoposide-induced apoptosis from 18% to 29% apoptosis (Fig. 1B). Herceptin treatment alone failed to increase apoptosis in the cell lines tested (Fig. 1C). Taken together, herceptin has the ability to increase etoposide-induced apoptosis in MCF-7 and MDA-MB-231 cells.

Herceptin treatment reduces ErbB2 and Mcl-1 expression in MDA-MB-231 cells. The mechanism for ErbB2-mediated cell survival is unclear. Expression of Bcl-2 family members regulate apoptosis (14, 18). MDA-MB-231 cells were treated with 1 or 10 μg/mL herceptin and mRNA levels of Bcl-2 family members were determined. We found that Bcl-x_L levels increased to 1.8-fold following 10 μg/mL herceptin treatment, whereas the proapoptotic family member Bax levels remained the same, and Mcl-1 mRNA levels decreased (Fig. 2A). Besides mRNA levels, Mcl-1, Bax, and Bcl-x_L protein levels were determined. Following treatment with 1 and 10 μg/mL herceptin, the amount of Mcl-1 was significantly reduced, but levels of Bax and Bcl-x_L seemed slightly increased (Fig. 2B). In addition, ErbB2 expression decreased following herceptin treatment (Fig. 2B), which indicates inactivation of ErbB2 signal transduction (19). This suggests that herceptin treatment reduces Mcl-1 protein expression compared with other Bcl-2 family members by inhibiting ErbB2 signaling.

Antisense against Mcl-1 renders MDA-MB-231 cells sensitive to etoposide-induced apoptosis. Antisense oligonucleotides against Bcl-2 have been shown to be effective at sensitizing cancer cells to apoptosis (20, 21). To determine whether reduced Mcl-1 expression will sensitize cells to etoposide-induced apoptosis, antisense oligonucleotides against Mcl-1 were transfected into MDA-MB-231 cells and the amount of etoposide-induced apoptosis was determined. Previously, we and other have shown that antisense oligonucleotides against Mcl-1 were specific for Mcl-1 and not other Bcl-2 family members (14, 22, 23). Etoposide treatment alone induced 18% apoptosis, whereas antisense, sense, and scrambled oligonucleotides against Mcl-1 failed to induce apoptosis. When etoposide was treated in combination with antisense against Mcl-1, the amount of apoptosis increased. In contrast, the combination of etoposide and sense or scrambled oligonucleotides failed to increase the amount of apoptosis (Fig. 3A). In addition, the antisense against Mcl-1 was transfected into MDA-MB-231 cells in combination with herceptin (1 μg/mL) and etoposide (1.5 mmol/L). The amount of apoptosis increased to 38% but did not give an additive effect (predicted additive effect would be 45%; Fig. 3B). To ensure that antisense against Mcl-1 was reducing Mcl-1 protein levels, cells were lysed and Western blotted for Mcl-1 in the presence or absence of antisense oligonucleotides. This indicated that antisense oligonucleotides decreased Mcl-1 expression (Fig. 3C). Taken together, this suggests that herceptin and antisense against Mcl-1 target similar antiapoptotic pathways.

Overexpression of Mcl-1 blocks etoposide-induced apoptosis. Decreased levels of Mcl-1 expression could explain the increase in the amount of etoposide-induced apoptosis following herceptin treatment. To determine whether overexpression of Mcl-1 blocks etoposide-induced apoptosis, we overexpressed Mcl-1 in the epithelial derived human embryonic kidney (HEK) 293 cells and treated the cells with etoposide. In cells overexpressing Mcl-1, the amount of etoposide-induced apoptosis was reduced compared to cells transfected with vector alone as indicated by changes in cellular morphology (Fig. 4A) and by chromatin condensation (Fig. 4B). Overexpression of Mcl-1, however, increased the amount of apoptosis in untreated cells (Fig. 4B). Cells were lysed and Western blotted for Mcl-1 to indicate overexpression of Mcl-1.

ErbB2 and Mcl-1 expression are associated in estrogen receptor–negative breast cancers. ErbB2 expression is often increased in cancer, especially in breast cancer (1). We obtained 29 estrogen receptor–negative tumors from the Manitoba Breast Tumor Bank and immunofluorescently stained for ErbB2 and Mcl-1. Scoring of ErbB2-positive tumors was conducted according to standard pathology practices (24, 25). We found that tumors that were positive for ErbB2 were all positive for Mcl-1 expression (Fig. 5A; Table 1). In tumors that were positive for Mcl-1 expression, only one showed low staining for ErbB2 (Table 1). These results were statistically significant using a χ² test (P < 0.00001). Furthermore, using Western blotting for ErbB2 and Mcl-1, tumors with high ErbB2 expression showed high Mcl-1 protein levels whereas tumors with low ErbB2 expression could express low or high Mcl-1 protein levels (Fig. 5B).

Exclusive ErbB2 expression in cells confers protection against taxol- and etoposide-induced apoptosis that is inhibited by herceptin treatment. To evaluate the role ErbB2 plays in inhibiting apoptosis in the absence of other EGF receptor family members, we used NIH3T3 cells that have been selected for low expression of EGF receptor family members (a kind gift from Dr. Hynes, Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland, ref. 15). These cells were either parental cells or parental cells stably transfected with human ErbB2 cDNA (NE2). The cells were then treated with a microtubule toxin taxol or etoposide. Untreated parental cells showed higher levels of apoptosis compared with NE2 cells (Fig. 6). On treatment of taxol over a dose range of 2.5 to 12.5 μmol/L, the amount of apoptotic cells increased to a maximum of 38% in parental cells after 48 hours. In contrast, the amount of apoptosis in NE2 cells following taxol treatment increased to only 15% after 48 hours (Fig. 6A). When etoposide (100 μmol/L) was added to parental cells, the amount of apoptosis increased to 40% at 48 hours. Similar to taxol, NE2 cells exposed to etoposide for 48 hours showed only 27% apoptosis (Fig. 6B). This indicates that expression of human ErbB2 decreases the amount of apoptosis in these cells following apoptotic stimuli.

To confirm that ErbB2 is conveying this resistance to apoptosis, parental and NE2 cells were exposed to etoposide alone, or in combination with increasing concentrations of herceptin (0.1-10 μg/mL). As the concentration of herceptin increased, the amount of apoptosis in NE2 cells increased following etoposide treatment until the level of apoptosis was similar to parental cells treated with etoposide (Fig. 6C). Herceptin treatment did not alter the amount of etoposide-induced apoptosis in parental cells due to the lack of ErbB2 expression. Similar results were found after taxol treatment with the addition of 1 μg/mL herceptin increasing taxol-induced apoptosis in NE2 cells from 15% to 24% (Fig. 6D). Herceptin alone failed to induce apoptosis in NE2 cells over a range of concentrations (Fig. 6E).

The expression levels of antiapoptotic Bcl-2 family members were also determined in parental and NE2 cells. We found that parental cells expressed detectable levels of Bcl-x_L and Bcl-2 but failed to express detectable levels of Mcl-1. In contrast, NE2 cells expressed all three Bcl-2 family members (Fig. 7A). To confirm that NE2 cells express ErbB2 and parental cells express low levels of ErbB2, the cells were lysed and Western blotted for ErbB2. As expected, NE2 cells expressed higher levels of ErbB2 than parental cells (Fig. 7A). Because herceptin increases etoposide-induced apoptosis, we determined whether herceptin treatment changes expression of Mcl-1 protein levels in NE2 cells. Using 10 μg/mL of herceptin, the level of Mcl-1 was significantly reduced at 24 hours, whereas 1 μg/mL herceptin did not reduce Mcl-1 protein levels (Fig. 7B). To determine whether lower concentrations of herceptin reduces Mcl-1 protein levels, NE2 cells were treated with 1 μg/mL of herceptin over a 72-hour time course. This resulted in decreased levels of Mcl-1 after 48 hours of herceptin treatment (Fig. 7C). This further illustrates that herceptin treatment reduces Mcl-1 expression and sensitizes cells to apoptosis when ErbB2 is expressed.

In many tumors, ErbB2 expression contributes to tumor progression and is a target for therapy (1, 2, 4, 5). ErbB2 expression has been correlated with increased drug resistance and aggressive disease (1). Inactivation of EGF signaling pathways using antibodies against ErbB2 or EGF receptors has been effective at increasing chemotherapy-induced apoptosis (5). Herceptin that targets ErbB2 is routinely used to treat breast cancer patients with increased ErbB2 expression (2). Herceptin is most effective when used in combination with standard chemotherapy (26). The increase in apoptosis associated with combined herceptin and chemotherapy has been attributed to inhibition of cell survival signaling pathways such as AKT signaling pathways (26, 27). Indeed, breast tumors with reduced PTEN expression are resistant to herceptin treatment due to increased AKT activation (27). This, however, only attributes the signaling pathways that block apoptosis but fails to account for herceptin-mediated activation of apoptotic pathways. For the first time, we have shown that herceptin reduces the expression of the antiapoptotic protein Mcl-1, and that elimination of Mcl-1 expression renders cells sensitive to chemotherapeutic drugs similar to herceptin treatment whereas overexpression of Mcl-1 inhibits apoptosis.

Members of the Bcl-2 family are one of the most important regulators of apoptotic potential (7, 8). Alterations in Bcl-2 family members are often found in cancers (8). Mcl-1 is overexpressed in many cancers (13, 28). In lymphoma, Mcl-1 is increased and contributes to tumor progression (29). In ovarian cancer, Mcl-1 expression correlates with poor prognosis (30). In chronic lymphocytic leukemia, Mcl-1 levels correlate with drug resistance and mutations in the Mcl-1 promoter seems to increase Mcl-1 expression (31–33). In invasive breast tumors, Mcl-1 expression levels failed to correlate with disease outcome (25), but we have now shown that Mcl-1 expression is increased by overexpression of ErbB2 in estrogen-negative tumors. Taken together, Mcl-1 seems to promote cell survival and its expression is elevated in cancer cells.

Besides cancer cells, Mcl-1 plays a critical role in hematopoietic stem cell survival. Mcl-1 is expressed at high levels in hematopoietic stem cells, B cells, and T cells (11, 34). Targeted deletion of Mcl-1 in these cell types leads to a loss in ear bone marrow progenitor cells, mature B lymphocytes, and T lymphocytes. Moreover, growth factors such as stem cell growth factor, interleukin 7, and interleukin 2 increases and/or maintains Mcl-1 expression levels in hematopoietic cells similar to cancer cells (11, 34). The importance of Mcl-1 in development was also shown in mice lacking Mcl-1 expression, which fail to develop past the blastocyst stage and fail to implant in utero, giving an embryonic-lethal phenotype (35). Mcl-1 seems to provide a survival response that is tightly regulated by growth factors in several different cell types.

Mcl-1 could be cleaved into a proapoptotic protein similar to Bcl-2 when cells undergo apoptosis or when Mcl-1 is overexpressed (12). Indeed, we have shown that overexpression of Mcl-1 increases the amount of apoptosis in untreated cells. However, we failed to detect cleavage of Mcl-1 in cells expressing endogenous Mcl-1 and in cells treated with herceptin. Furthermore, etoposide failed to increase the amount of apoptosis in Mcl-1-overexpressing cells. This indicates that Mcl-1 is acting as an antiapoptotic protein in cancer cells.

Growth factors and their corresponding receptors have been shown to promote cell survival in cancer cells (36). Activation of EGF receptor family members leads to up-regulation of Mcl-1 (14, 37). This up-regulation has been associated with protection against apoptosis following apoptotic stimulation. Indeed, chemotherapeutic drug–induced apoptosis is effectively blocked by overexpression of EGF receptor family members and Mcl-1 (14, 37–39). In chronic lymphocytic leukemia, Mcl-1 protein levels decreased following treatment with genotoxic agents (33). We have previously shown that EGF binding to its receptors leads to increased Mcl-1 expression preventing death receptor–induced apoptosis (14). ErbB2 overexpression is also effective at blocking both death receptor– as well as chemotherapeutic drug–induced apoptosis (19). This supports our results that herceptin treatment reduces Mcl-1 levels and renders cells sensitive to chemotherapy.

Regulation of Mcl-1 proteins levels could occur in several ways. We and others have shown that EGF activates the phosphoinositide-3-kinase/AKT signaling pathway, leading to activation of the transcription factor nuclear factor κB and up-regulation of Mcl-1 (14, 40). In ErbB2-positive breast tumors, nuclear factor κB is preferentially activated (41), corresponding with our findings that Mcl-1 expression is associated with ErbB2-positive tumors. Furthermore, the stability of Mcl-1 mRNA and proteins is shorter than other Bcl-2 family members (13). This implies that subtle alterations in signaling pathways such as ErbB2 overexpression could change the expression levels for Mcl-1 by altering the stability of either mRNA or proteins without affecting other Bcl-2 family members. In NE2 cells, Bcl-2 and Bcl-x_L protein levels did not change compared with parental cells but Mcl-1 expression increased. In MDA-MB-231 cells, it seems that Mcl-1 mRNA levels decrease, suggesting alterations in mRNA stability or transcription. It is still possible that ErbB2 signaling could be stabilizing the protein because stabilization of p21^CIP1 by EGF receptors has been shown (42). In chronic lymphocytic leukemia, Mcl-1 proteins levels decreased following treatment with genotoxic agents, but its mRNA levels remain unchanged (33). In addition, Mcl-1 is unique in the Bcl-2 family in having a role in cell cycle control, and by binding to proliferating cell nuclear antigens, could prevent the passage of cells through S phase (26). This is similar to herceptin being implicated in retarding tumor growth (43). Our results suggest that sensitization of breast cancer cells to apoptosis following herceptin treatment is due to decreased Mcl-1 protein expression. The regulation of Mcl-1-decreased expression of Mcl-1 by ErbB2 will be the focus for future investigations.

Targeting Bcl-2 family members with therapeutics could be a good strategy to sensitize cancer cells to undergo apoptosis (21). Antisense oligonucleotides against Bcl-2 have been shown to be effective at inducing apoptosis in several leukemia cells that have high Bcl-2 expression (21). This drug is currently in clinical trials for several leukemias and has shown some efficacy (44). The development of small molecule inhibitors of Bcl-2 or Bcl-x_L interactions with proapoptotic Bcl-2 family members such as Bax has been successfully shown to induce apoptosis in a variety of cancer cell lines (45–47). These molecules bind to the BH3 cleft found in Bcl-2 and Bcl-x_L preventing binding of Bax or Bak. It has been shown that these BH3-like molecules could also interfere with Mcl-1 binding to proapoptotic Bcl-2 family members (48–50). Similarly, antisense oligonucleotides against Mcl-1 have been effective at inducing apoptosis in several cell lines and, in combination with chemotherapy, enhances apoptotic responses in cell lines and in xenografted tumors (20, 22, 23). Using small interfering RNA to target Mcl-1 expression was also effective at inducing and enhancing apoptosis in cancer cells (51). Our data suggests that treatment with antisense oligonucleotides against Mcl-1 might render cells sensitive to chemotherapy and reduce drug resistance but addition of herceptin will not enhance this apoptotic response. Thus, overexpression of ErbB2 could increase drug resistance and cell survival by inducing Mcl-1 expression. Targeting Mcl-1 expression could be an effective treatment in herceptin-resistant ErbB2-overexpressing breast tumors.

We thank Dr. Hynes for the kind gift of reagents, Drs. Watson and Leygue for their advice and help with the preparation of this manuscript, and the Manitoba Breast Tumor Bank for supplying the tumor samples.