Is Caspase-8 a Neuroendocrine Lung Tumor Supressor?

Apoptosis, a universal genetic program of cell death in higher eukaryotes, is a basic process involved in cellular development and differentiation.¹ Apoptosis may be essential for the prevention of tumor formation, and its deregulation is widely believed to be involved in pathogenesis of many human diseases, including cancer.² Apoptosis is caused by activation of intracellular proteases, known as caspases, which are responsible directly or indirectly for the morphological and biochemical events that characterize apoptotic cells. A number of caspase regulators have been characterized in their response to environmental or therapeutic stimuli.³ Cell death ligands, such as CD95L and Apo2L/TRAIL trigger apoptosis through assembly at the plasma membrane of a protein complex which constitutes the death-inducing signaling complex (DISC).⁴ The CD95, DR4, and DR5 receptors for these ligands recruit the adaptor molecule FADD/Mort1, which in turn brings into the complex the apoptosis intitiator caspases 8 and/or 10.⁵ The studies of Shivapurkar et al, appearing in this issue of the journal, implicate caspase 8 (CASP8) gene expression in lung cancer. A first hint of a possible involvement of CASP8 in lung cancer was revealed following its localization to chromosome 2q33 and identification of a homozygous deletion at this locus in a SCLC cell line.⁶ The data of Shivapurkar et al indicate that CASP8 gene expression was lost in majority (79%) of SCLC but no NSCLC tumor cell lines. This discovery is very important as it highlights one of the very few absolute differences between molecular markers of SCLC and NSCLC. The loss of CASP8 gene expression corresponds to a similar loss of protein and enzymatic activity and occurrs through methylation of the promoter in 59% of the cell lines examined which lacked CASP8 expression. With rare exceptions, cell lines harboring a methylated allele lacked the corresponding unmethylated allele, indicating either biallelic methylation or allelic loss of the unmethylated allele. This proportion is significant, as the authors have shown previously that losses in lung cancer cell lines over 30% were abnormal and indicators of the location of tumor suppressor genes.⁷ The mechanism responsible for inactivation of gene expression in the remaining cell lines remains to be explored. Using fresh tissue and using 5 polymorphic markers, Shivapurkar et al finally show that most SCLC and many bronchial carcinoids had CASP8 promoter methylation, which would corespond to 58% SCLC and 30% bronchial carcinoids lacking CASP8 expression.

DNA methylation of promoter and regulatory sequences in the 5′ untranslated region (UTR) may be the mechanism of inactivation of many tumor suppressor genes, including the cell cycle regulators p15, p16, ARF, p73, and pRb.⁸ Remarkably, Shivapurkar et al demonstrate that treatment of SCLC cell lines lacking CASP8 expression with the demethylating agent 5-aza-2′-deoxycytidine (5-dAzaC), restored gene expression in 75% of the SCLC cell lines tested. Similarly, expression of CASP8 silenced in neuroblastomas could also be restored.⁹ In a variety of other tumor cells including Ewing, neuroblastoma, malignant brain, and melanomas, caspase-8 expression acts as a key determinant of sensitivity for apoptosis induced by death-inducing ligands or cytotoxic drugs. Treatment with 5-dAzaC reversed hypermethylation of caspase-8 resulting in restoration of caspase-8 expression and recruitment and activation of caspase-8 at the CD95 DISC upon receptor cross-linking, thereby sensitizing for death receptor-, and drug-induced apoptosis. Inhibition of caspase-8 activity also inhibited apoptosis sensitization by 5-dAzaC.¹⁰ It has yet to be determined whether restored CASP8 levels could lead to increased CASP8 activity in lung tumors in response to apoptotic stimuli.

A previous study, with fewer cell lines, has indicated that SCLC was characterized by loss of expression of CASP8, in addition to which also of CASP10, CASP1, and CASP4.¹¹ The possibility that other caspases might be inactivated in SCLC is not explored in this study. Particulary, loss of CASP10 would be of interest as its inactivation seem to be more widespread than that of CASP8 in human cancers.⁵ Moreover, the CASP8 and CASP10 loci are linked, being localized to chromosome 2q33, only 40 kb apart.¹² Indeed, a recent report involving also the authors’ laboratory does indeed suggest that CASP10 expression may be absent in lung and breast cancer cell lines, even though RNA is expressed, indicating yet another mechanism for inactivation of the apoptotic apparatus during tumor development.¹³

Paradoxically, SCLC tumors and cell lines which have lost CASP8 expression have been reported to have high spontaneous apoptosis rates.¹¹ At the same time, they also have high proliferation rates,¹¹ consistent with the notion that often these processes are inter-related.¹⁴ In fact, in neuroblastomas, where a similar CASP8 inactivation has been reported, it has been associated with MYCN amplification, leading to increased proliferation.⁹ Interestingly, defects in cell cycle control mechanisms involving primarily the Rb pathway are 100% inactivated in many cancers, including that of the lung.¹⁵ Studies over the last decade have indicated that in most human cancers, including that of the lung, cells sustain mutations that affect the functions of Rb and p53, either by disabling these genes directly or by targeting genes that act epistatically to prevent their proper function.¹⁶ Indeed, all SCLCs and many large cell neuroendocrine carcinomas (LCNECs) have abnormalities in the p16:RB pathway.¹⁷ Whether cell cycle abnormalities and CASP8 and/or 10 expression are both necessary for lung tumor development remains to be determined.

In conclusion, the paper of Shivapurkar et al provides critical evidence for the lack of CASP8 expression in a subset of both high grade (SCLC) and low grade (carcinoid) neuroendocrine lung tumors, but not NSCLC, which usually lack neuroendocrine characteristics. Therefore CASP8 may function as a tumor suppressor in neuroendocrine lung tumors, similarly to neuroblastomas, also of neuroendocrine origin. So could CASP8 be a tumor supressor for lung cancer? Indeed, it has been shown that CASP8 acts as tumor suppressor gene in neuroblastoma⁹ and it may have a similar function in other tumors, such as Ewing, malignant brain, and melanomas.¹⁰ Similar to demethylation, introduction of caspase-8 by gene transfer sensitizes for apoptosis induction, indicating that re-expression of caspase-8, by demethylation or caspase-8 gene transfer, might be an effective strategy to restore sensitivity for chemotherapy- or death receptor-induced apoptosis in various tumors in vivo,¹⁰ including that of the lung. The paper of Shivapurkar et al, taken together with a recent report implicating mutations in the DR4 receptor for Apo2L/TRAIL in NSCLC¹⁸ indicate that both the neuroendocrine and non-neuroendocrine lung tumors might have the death ligand-receptor pathway to apoptosis inactivated, but through different mechanisms.