Assessment of TRAIL sensitisation by IAP antagonist TL32711 in malignant melanoma and development of a framework for response prediction

Abstract

Despite the availability of different therapeutic options, large numbers of melanoma patients still do not respond to treatment or experience disease recurrence. Therefore, there is a need to develop novel therapeutics and identify those patients who would benefit most from such therapies. Inducing apoptosis with the cytokine tumour necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) is a promising way of eliminating cancer cells as it solely activates cell death in malignant tissue. Due to this specificity, recombinant TRAIL and agonistic antibodies against TRAIL receptors 1 and 2 have been tested in clinical trials, however only with limited success. Besides low stability in serum and low efficacy in crosslinking receptors, TRAIL-based therapy also failed due to the intrinsic resistance of cancer cells caused by the upregulation of anti-apoptotic proteins. Therefore, in this study, a novel 2nd generation TRAIL receptor agonist, IZI1551, with improved stability and increased valency was used to investigate apoptosis induction in melanoma cell lines. In addition, to sensitise cells to TRAIL-induced apoptosis, IZI1551 was combined with Birinapant, a SMAC mimetic designed to degrade anti-apoptotic IAP proteins. The combination treatment was tested in a large melanoma cell line panel, covering different mutation status and different disease stages. Furthermore, results of these experiments together with data about the expression of apoptosis-related proteins were used to predict the responsiveness of melanoma to the treatment. Across the cell line panel, high heterogeneity in responsiveness was observed, ranging from complete resistance against the combination treatment with TRAIL and Birinapant to pronounced synergies between the two drugs. More specifically, the combination treatment induced apoptosis in 12/16 melanoma cell lines, 4/5 3D spheroids, and 3/5 patient-derived melanoma cell lines. It was shown in this work that Birinapant efficiently depleted cIAP1 in the first hour of treatment and that depletion lasted for at least 24 h. This, inconsistency with current literature available, most likely caused synergistic cell death in melanoma cell lines when treated with Birinapant in combination with IZI1551. Previous studies showed that the responsiveness of the cell lines to TRAIL could not be concluded from the expression of individual apoptosis proteins. Therefore, in this study, a data-driven modelling strategy was devised with the aim to classify cells according to their responsiveness to the combination. treatment with IZI1551 and Birinapant. For that, the basal expression of 19 cell death regulators, known to be relevant for the responsiveness to IZI1551 was determined and included in the model. Indeed, by using multivariate statistics, responsive cell lines could be separated from low-responding cell lines with 87.5% accuracy while responsiveness predictions achieved 81.25% accuracy. The predictive capacity of the model was validated for 3D growth conditions (80% accuracy, n=5), patient-derived melanoma cell lines (100%, n=5) and melanoma cell lines grown as xenografts (100%, n=4). Taken together, the combination treatment with IZI1551 and Birinapant induced apoptosis in the majority of melanoma cell lines, 3D melanoma spheroids and patient-derived melanoma cell lines. Moreover, responders and non-responders were successfully identified based on their apoptosis protein expression profiles by using data-driven modelling. In conclusion, the capability to predict responsiveness to combinations of TRAIL receptor agonists and SMAC mimetics might provide an avenue for the future development of personalized treatment strategies based on these targeted therapeutics.