Arginine deprivation therapy induces apoptotic cell death in melanoma brain metastasis

Abstract Aims The development of melanoma brain metastasis (MBM) occurs in ~50% of metastatic melanoma cases, and significantly worsens prognosis to a median survival of 12.8 months. Melanoma is often reported as an arginine auxotroph due to transcriptional silencing of argininosuccinate synthase 1 (ASS1). Arginine deiminase (ADI) is a non-mammalian enzyme which depletes blood arginine by converting it to citrulline and ammonia, and in its pegylated form ADI shows clinical efficacy in the treatment of a number of cancers via exploiting tumour arginine auxotrophy, resulting in targeted arginine deprivation of tumour cells. While cutaneous melanoma is the prototype cancer for this therapy, studies to date have excluded central nervous system metastasis. We have demonstrated that patient derived primary MBM models are sensitive to arginine deprivation in vitro, confirmed suitable clinical biomarkers of sensitivity, and established the mechanism of tumour cell specific cytotoxicity. Method Patient derived primary cultures of MBM were established and subject to treatment with arginine deprivation. Gene expression and methylation analysis was examined by RT-qPCR, western blot, Illumina mRNA sequencing and Illumina methylated DNA immunoprecipitation-sequencing (MeDIP-seq) on ADI treated and untreated samples. Cell death, cytotoxicity induction and caspase-3 and-7 recruitment was analysed using an Incucyte S3 live-cell imager, by fluorescently labelling cells with Incucyte Cytolight Red Rapid dye, Cytotox Green dye and Caspase-3/7 Green dye, and imaging cells every 2 hours over the course of 2 weeks. 3D spheroid growth and invasion was measured by culturing cells as tumour spheroids before treating with ADI, and imaging spheroids every 2 hours for 2 weeks using an Incucyte S3 live-cell imager. Nuclear leakage and mitochondrial morphology was observed by fluorescently staining treated and untreated cells with DAPI and MitoTracker Red, and imaging on a Leica DMi8 confocal microscope. Results Primary MBMs differentially express ASS1 at substantially lower levels than non-cancerous melanocytes, however some models are capable of upregulating ASS1 following confrontation with arginine deprivation. Despite this, long-term sensitivity of primary MBMs to arginine deprivation was observed in both 2D and 3D models. In addition, arginine deprivation was seen to inhibit MBM invasion in a 3D model – an important feature in MBM pathogenesis. Initially, autophagy was induced in arginine deprived MBM, however in all models the induction of cytotoxicity correlated with recruitment of caspase-3 and -7, and intrinsic apoptotic cell death confirmed. Nuclear leakage, and eventually complete nuclear destruction was observed, in addition to mitochondrial fragmentation. Conclusion Arginine deprivation is highly effective in reducing 2D and 3D MBM growth, as well as limiting invasion. While apoptotic cell death was observed in all models, the initial induction of autophagy could pose threat of resistance development in a clinical setting, and so combinational therapies with autophagic inhibitors and/or additional apoptotic inducers should be investigated. It is unclear whether nuclear leakage and mitochondrial degradation are the cause or product of apoptosis. Considering the strong clinical evidence for the use of arginine deprivation in non-CNS metastatic melanoma and the results of this study, arginine deprivation is a highly suitable treatment for pre-surgical MBM to limit invasion and increase resection, and for post-surgical continuation.