Targeting bcl-2 in Small-Cell Lung Cancer
ABSTRACT & COMMENTARY
Synopsis: Small-cell lung cancer cell lines express high levels of bcl-2 mRNA and are resistant to chemotherapy-induced apoptosis. Culture of small-cell lung cancer cell lines with specific antisense oligonucleotides leads to reduced expression of bcl-2 and subsequent reduction in the viability of cancer cells and the induction of apoptosis. In vivo administration of antisense oligonucleotides targeted to bcl-2 may have direct anti-tumor effects or may potentiate the efficacy of chemotherapy.
Source: Ziegler A, et al. J Natl Cancer Inst 1997;89: 1027-1036.
The bcl-2 gene, which was first identified in a follicular lymphoma carrying a t(14;18) translocation, functions as a suppressor of apoptotic cell death. The Bcl-2 protein prevents activation of a protease cascade that ends in cell death and cooperates with Myc oncoproteins to transform cells. In addition, the overexpression of bcl-2 appears to confer resistance to chemotherapy (cisplatin and etoposide) and radiation in neuroblastoma and leukemia cell lines. In leukemia cells, interference with apoptosis can be reversed by downregulating bcl-2 expression.
Since 90% of small-cell lung cancers and small-cell lung cancer cell lines overexpress bcl-2, and because transfection of small-cell lung cancer cells with the gene for bcl-2 resulted in increased resistance to chemotherapy-induced apoptosis, Ziegler and colleagues hypothesized that bcl-2 must be critical for inhibiting apoptosis in small-cell lung cancer cells and, therefore, may contribute to the development of drug resistance. Inhibiting bcl-2 expression in small-cell lung cancer cells with antisense oligodeoxynucleotides (ODNs) to bcl-2 messenger RNA could lead to the reversal of this drug-resistant phenotype.
To test their hypothesis, Ziegler et al prepared 13 ODNs, all 20 nucleotide residues long, to inhibit bcl-2 mRNA expression. Four small-cell lung cancer cell lines were treated with the various ODNs and tumor cell cytotoxicity was measured by a colorimetric assay, and apoptosis was identified by light scattering and the presence of DNA fragmentation. Northern and Western blot analysis was performed to quantify the expression of bcl-2 mRNA and Bcl-2 protein, respectively. All four cell lines expressed high levels of bcl-2 mRNA, but without much correlation with the levels of Bcl-2 protein. Two cell lines required long periods of incubation for low levels of Bcl-2 protein to be detected. A cell line with intermediate levels of Bcl-2 protein exprerssion was used to initiate further studies.
Although previous data suggested that ODNs directed against the translation initiation portion of mRNA were likely to be most cytotoxic, of the 13 ODNs tested, ODN 2009, which targets a midportion of the mRNA, was the most cytotoxic ODN leading to a reduction in viability to 10%. Western blots showed that ODN 2009 also reduced Bcl-2 protein levels to 18% at 24 hours and to 48% at 96 hours. Appropriate controls using ODNs with a scrambled sequence demonstrated that sequence-specific ODNs were needed to observe these effects.
Ziegler et al also examined the effects of ODNs on apoptosis by incubating one of the cell lines (SW2) with several concentrations of ODN 2009. A dose-dependent increase in apoptosis was demonstrated. Incubation of the three other cell lines with ODN 2009 also resulted in the variable induction of apoptosis that showed good correlation with the cell lines’ basal Bcl-2 protein levels (r = -0.9964; P = 0.003).
Lung cancer is the largest cause of cancer deaths in the West, and small-cell lung cancer comprises about 25% of these cases. Current cytotoxic treatments, with or without radiation therapy, yield a two-year survival of approximately 5%. Novel genetic mechanisms that may contribute to drug resistance have been identified (e.g., alterations in drug transport and defects that prevent apoptotic cell death after cytotoxic damage); the ability to overcome these genetic mechanisms may improve the therapy of this disease. This article, others like it, and several presentations at this year’s ASCO meeting may herald a fourth wave of antineoplastic drug therapy after endocrine, cytotoxic, and immunological approaches. Many common tumor types express novel proteins that are the products of oncogenes or defective anti-oncogenes (tumor suppressor genes). Some of these proteins are fairly specific for certain neoplasms such as bcr-abl in CML, HER-2/neu in breast and ovarian cancer, and VHL in renal cell carcinoma. Others such as Bcl-2 and defective p53 are found in a range of tumors. The attraction of these proteins as targets lies in their role as active participants in driving neoplasia as well as their relatively selective expression in the neoplastic cell.
The Ziegler article details early investigations in attacking bcl-2 in small-cell lung cancer. The protein product of this gene, Bcl-2, inhibits apoptosis occurring spontaneously as well as following exposure to cytotoxic drugs or ionizing radiation. Although small-cell lung cancer lines vary in their expression of Bcl-2 protein, the overwhelming majority appear to express high levels. Since Bcl-2 protein has been shown to protect against chemotherapy-induced apoptosis in several neoplastic cell lines, it is easy to hypothesize that cytotoxic chemotherapy results in destruction of the cells expressing low levels of Bcl-2 and that the commonly seen rapid recurrence of this tumor reflects growth of relatively Bcl-2 enriched clones. Bcl-2 may be an interesting target for intervention. ODNs targeting a variety of portions of the mRNA molecule are effective in different cell lines. Ziegler and coworkers used 13 ODNs to target various portions of the molecule. ODN 2009, which targets a midportion of the coding sequence, proved to be most cytotoxic. ODN 2009 produced the most cytotoxicity as well as most durable suppression of Bcl-2. The ODN 2009 effect showed good dose-related effects. ODN 2009 was cytotoxic to cell lines that had lower levels of Bcl-2 protein, but this killing effect correlated with the Bcl-2 levels. The correlation was disproportional, which indicates that there may be a complex interplay of pro- and anti-apoptotic molecules that can be targets for therapeutic manipulation.
Missing from the Ziegler et al paper is any indication of in vivo efficacy. Phase I trials of antisense oligonucleotides are being performed. Results so far are underwhelming. ODNs to p53 are being investigated in patients with leukemia, and Bcl-2 antisense ODNs have been administered to lymphoma patients. No study has demonstrated dose-limiting toxicity. There is little information on the pharmacology of these agents in people. It is not even clear whether the antisense ODNs have actually reduced the levels of expression of the targeted messages. However, occasional patients on the phase I studies have been said to experience tumor regressions, and, as our colleague, Dan Von Hoff, has pointed out, every drug that has shown responses in phase I studies has gone on to have useful antitumor activity. So, there may be reason to be hopeful. Yet it seems unlikely that every target will be amenable to manipulation.
This article and an accompanying editorial (Reed JC. Promise and problems of Bcl-2 antisense therapy. J Natl Cancer Inst 1997;89:988-990) point out some of the problems that need attention as these compounds enter clinical trials. ODN selection is important. The ODNs are complimentary to relatively small portions of the bcl-2 mRNA. A priori, it was thought that a translation initiation site binding ODN would be most cytotoxic, but a molecule that bound a midportion of the coding sequence was found to be most cytotoxic. Although the authors screened 13 ODNs, there certainly may be others that are equally, or even more, effective. It is not known whether combinations of ODNs may be synergistic. Delivery of the ODNs to patients could present problems. For the in vivo experiments, the ODNs were dissolved in a solvent. Formulation issues will need to be addressed for clinical trials. In patients, ODNs may need to be given in vectors such as viruses.
The prospect of selectively antagonizing the expression of proteins involved in the maintenance of the neoplastic state is attractive. However, we have learned from hard experience that hitting a single target is rarely sufficient to kill a tumor. Perhaps the most enticing aspect of ODN work is the possibility that ODNs blocking drug resistance or apoptosis resistance mechanisms may enhance the antitumor effects of other agents.