Post-translational modifications of c-FLIP: opportunities for therapeutic intervention

  • Joel Stephen Riley

Student thesis: Doctoral ThesisDoctor of Philosophy

Abstract

FLIP is an anti-apoptotic protein found overexpressed in a wide variety of cancers and is associated with drug resistance and poor prognosis. It acts by inhibiting the membrane-proximal steps of death receptor-mediated apoptosis at the death inducing signalling complex (DISC), but also has roles in secondary complexes. Both the long and short splice forms of FLIP contain two tandem death effector domains, homologous to procaspase-8. FLIP preferentially binds to FADD via its DEDs, inhibiting the recruitment and processing of procaspase-8 to its active form. Normal cells are resistant to loss of FLIP, but cancer cells are not. Targeting FLIP in cancer is an attractive therapeutic approach, though due to its lack of enzymatic activity strategies have remained elusive.

  A tissue microarray study performed on 184 samples from non-small cell lung cancer patients revealed that FLIP and procaspase-8 are both overexpressed in nuclear and cytoplasmic compartments of tumoural tissue compared to surrounding stromal tissue. This was observed in both adenocarcinoma and squamous histologies and analyses showed that high FLIP correlates with high procaspase-8. Survival analyses High cytoplasmic FLIP was associated also with poor survival.

  Pan-HDAC inhibitors were identified as modulators of early FLIP down- regulation in NSCLC cell lines. Treatment with HDACi induced apoptosis in a time- and dose-dependent manner concomitant with activation of caspases. However, normal cells were unaffected by treatment with HDACi, suggesting these agents might provide a good therapeutic window. Silencing procaspase-8 rescued the apoptotic phenotype observed with HDACi treatment, indicating that the extrinsic apoptotic pathway was critical in mediating apoptosis induced by HDACi. Further investigation showed that FLIPs overexpression and death receptors 4 and 5 silencing hampered HDACi-induced apoptosis.

  NSCLC is known to be a highly drug resistant disease. Patients are often resistant to current treatment options such as cisplatin so combining these with HDACi may provide clinical benefit. Indeed, we found that cotreatment of vorinostat and cisplatin synergistically enhanced apoptosis and reduced long-term clonogenic survival. Furthermore, apoptosis arising from vorinostat and cisplatin co-treatment was FLIPs- and caspase-8-dependent.

TRAIL is an anti-cancer agent that activates the extrinsic apoptotic cascade through the death receptors. Disappointingly it has failed to show significant patient benefit. A major mechanism of drug resistance to TRAIL is overexpression of FLIP. Having already shown that HDACi down-regulate FLIP at early time-points, we pre-treated NSCLC cells with vorinostat prior to treatment with TRAIL. This combination synergistically induced apoptosis, reduced long-term clonogenic survival and was dependent on FLIP and caspase-8.a time- and dose-dependent manner concomitant with activation of caspases. However, normal cells were unaffected by treatment with HDACi, suggesting these agents might provide a good therapeutic window. Silencing procaspase-8 rescued the apoptotic phenotype observed with HDACi treatment, indicating that the extrinsic apoptotic pathway was critical in mediating apoptosis induced by HDACi. Further investigation showed that FLIPs overexpression and death receptors 4 and 5 silencing hampered HDACi-induced apoptosis. NSCLC is known to be a highly drug resistant disease. Patients are often resistant to current treatment options such as cisplatin so combining these with HDACi may provide clinical benefit. Indeed, we found that cotreatment of vorinostat and cisplatin synergistically enhanced apoptosis and reduced long-term clonogenic survival. Furthermore, apoptosis arising from vorinostat and cisplatin co-treatment was FLIPs- and caspase-8-dependent.TRAIL is an anti-cancer agent that activates the extrinsic apoptotic cascade through the death receptors. Disappointingly it has failed to show significant patient benefit. A major mechanism of drug resistance to TRAIL is overexpression of FLIP. Having already shown that HDACi down-regulate FLIP at early time-points, we pre-treated NSCLC cells with vorinostat prior to treatment with TRAIL. This combination synergistically induced apoptosis, reduced long-term clonogenic survival and was dependent on FLIP and caspase-8.

  Vorinostat is a pan-HDACi with significant side-effects a poor half-life in humans. HDACi with greater selectivity are being developed for clinical use. One example is entinostat which has been shown to selectively inhibit
HDACs 1,2 and 3. We found that treatment with entinostat also downregulated FLIP and combination treatments with cisplatin or TRAIL were in most cases equally or more effective than vorinostat. Vorinostat-treatment induced the ubiquitination of both splice forms of FLIP and FLIP'S down-regulation was blocked with pre-treatment of the proteasome inhibitor MG-132. Mutants of FLIPs with ubiquitin-acceptor lysines mutated to arginine were generated and confirmed as ubiquitination sites in adherent, epithelial HCT116 colorectal cancer cells and 293T cells. Moreover, mutation of these lysines extended FLIP'S half-life in cells.

  Silencing of caspase-8 was found to increase FLIP'S ubiquitination at the DISC which was mediated through Lysl92 and Lysl95. Conversely, silencing of FLIP increased caspase-8's ubiquitination at the DISC. Plowever, cells stably overexpressing ubiquitin mutants of FLIPs conferred similar levels of resistance to TRAIL as wild-type.
  A handful of studies have identified ITCH as an E3 ligase which regulates FLIP'S ubiquitination. However, in our hands, silencing of ITCH did not appear to be important. No deubiquitinating enzymes (DUBs) have been associated with FLIP. To resolve this, we performed two screens: one to identify novel DUBs which may affect FLIP'S ubiquitination and a mediumthroughput luciferase-based screen to identify potential E3 ligases which alter FLIP'S stability. CSN5, a component of the COP9 signalosome was identified from a DUB screen for FLIPl, but had also appeared on a yeast-2- hyrbid screen for FLIPl from our laboratory. It was found to interact with FLIP in a ubiquitination-dependent manner and when silenced dramatically reduced FLIP'S ubiquitination. Furthermore, abrogation of CSN5 sensitised cells to TRAIL treatment.
  The ubiquitin-editing enzyme A20 and the DUB CYLD were hits chosen to be validated from a ubiquitin family esiRNA screen. Again, these were found to interact with FLIP by immunoprecipitation in a FLIP-ubiquitinationdependent manner and reduced FLIP'S ubiquitination when silenced.

  These studies provide a pre-clinical rationale for the use of HDACi, particularly when combined with TRAIL or cisplatin. Work to further understand the biology surrounding FLIP'S ubiquitination and stability may lead to a therapeutic strategy aimed at overcoming chemotherapy resistance by targeting FLIP'S expression.
Date of AwardSep 2013
LanguageEnglish
Awarding Institution
  • Queen's University Belfast

Cite this