Immunotherapy Bridge 2020 and Melanoma Bridge 2020: meeting abstracts Characterization of the metastatic behaviour and gene expres‐ sion profile (RNAseq) of different melanoma cell lines: a compre‐ hensive in vivo model.

Background: Metastasis is the major cause of death in malignant melanoma. Several factors, including clinical-pathological and tumor biological features may restrain prognosis. Molecular mechanisms regulating melanoma progression and metastasis have been partially discovered, and thus we developed an in vivo model of metastatic melanoma to investigate potential genes implicated in these events. Materials and methods: To evaluate the in vivo metastatic activity of five different melanoma cell lines (LCP, LCM, WM266, SK-Mel28 and A375), we completed intra-cardiac (ic) injection of 1 × 106 luminescent cells or PBS as control in three NOD-SCID mice per cell line. After 3 weeks, mice were studies by in vivo bioluminescence imaging (IVIS Lumina LT) to measure the mean radiance (p/sec/cm2/sr), that we arbi- trarily considered as a surrogate of the total metastatic tumor burden (t-MTB). Animals were euthanized at the humane endpoints achieve- ment and underwent X-ray evaluation for bone metastasis detec- tion. The Kaplan-Meyer curves described the time to sacrifice from ic injection. Mouse necropsy identified metastatic organs that were explanted and processed for both histology and gene expression anal- ysis. Melanoma cell lines were profiled for gene expression (RNAseq) of 118 genes notably involved in cancer progression and metastasis. Quantitative RT-PCR (qRT-PCR) explored the expression of a restrict number of genes on formalin-fixed paraffin-embedded (FFPE) meta- static samples from euthanized animals. Results: All melanoma cell lines demonstrated a metastatic behaviour following ic injection (Fig. 1) with a variable attitude to produce bone and visceral metastasis (Table 1). Mice injected with LCP, LCM and WM266 showed a lower t-MTB and increased survival (Fig. 2) as com- pared to A375 and SK-Mel28. Thus, we arbitrarily defined LCP, LCM and WM-266 cell lines ‘poorly metastatic’ (group A), while A375 and SK-Mel28 ‘highly metastatic’ (group B). The principal component analysis and the unsupervised hierarchical clustering (118 gene transcriptome heatmap) revealed similar gene expression profiling among cell lines grouped for the metastatic atti- tude (Fig. 3). The gene expression analysis (Fig. 4) performed on FFPE samples identified five deregulated genes (WNT5A, COL6A3, PTHLH, SOX9 and SERPINE1) between A and B. Conclusions: We describe the metastatic capacity of five melanoma cell lines. Gene expression profiling revealed the activation of five genes as putatively responsible for the high aggressiveness of A375 and SK-Mel28 cells. These results suggest to investigate these genes in a clinical setting and their possible application as druggable target for future therapeutic strategies.