BRAF fusions identified in melanomas have variable treatment responses and phenotypes.

Oncogenic BRAF fusions have emerged as an alternate mechanism for BRAF activation in melanomas and other cancers. A number of BRAF fusions with different 5' gene partners and BRAF exon breakpoints have been described, but the effects of different partners and breakpoints on cancer phenotypes and treatment responses has not been well characterized. Targeted RNA sequencing was used to screen 60 melanoma patient-derived xenograft (PDX) models for BRAF fusions. We identified three unique BRAF fusions, including a novel SEPT3-BRAF fusion, occurring in four tumors (4/60, 6.7%), all of which were "pan-negative" (lacking other common mutations) (4/18, 22.2%). The BRAF fusion PDX models showed variable growth rates and responses to MAPK inhibitors in vivo. Overexpression of BRAF fusions identified in our study, as well as other BRAF fusions previously identified in melanomas, resulted in a high degree of variability in 2D proliferation and 3D invasion between the different fusions. While exogenously expressed BRAF fusions all responded to MAPK inhibition in vitro, we observed potential differences in signaling and feedback mechanisms. In summary, BRAF fusions are actionable therapeutic targets, however there are significant differences in phenotypes, treatment responses, and signaling which may be clinically relevant.

BH3 mimetics induce apoptosis independent of DRP-1 in melanoma.

Despite the recent advancement in treating melanoma, options are still limited for patients without BRAF mutations or in relapse from current treatments. BH3 mimetics against members of the BCL-2 family have gained excitement with the recent success in hematological malignancies. However, single drug BH3 mimetic therapy in melanoma has limited effectiveness due to escape by the anti-apoptotic protein MCL-1 and/or survival of melanoma-initiating cells (MICs). We tested the efficacy of the BH3 mimetic combination of A-1210477 (an MCL-1 inhibitor) and ABT-263 (a BCL-2/BCL-XL/BCL-W inhibitor) in killing melanoma, especially MICs. We also sought to better define Dynamin-Related Protein 1 (DRP-1)'s role in melanoma; DRP-1 is known to interact with members of the BCL-2 family and is a possible therapeutic target for melanoma treatment. We used multiple assays (cell viability, apoptosis, bright field, immunoblot, and sphere formation), as well as the CRISPR/Cas9 genome-editing techniques. For clinical relevance, we employed patient samples of different mutation status, including some relapsed from current treatments such as anti-PD-1 immunotherapy. We found the BH3 mimetic combination kill both the MICs and non-MICs (bulk of melanoma) in all cell lines and patient samples irrespective of the mutation status or relapsed state (p < 0.05). Unexpectedly, the major pro-apoptotic proteins, NOXA and BIM, are not necessary for the combination-induced cell death. Furthermore, the combination impedes the activation of DRP-1, and inhibition of DRP-1 further enhances apoptosis (p < 0.05). DRP-1 effects in melanoma differ from those seen in other cancer cells. These results provide new insights into BCL-2 family's regulation of the apoptotic pathway in melanoma, and suggest that inhibiting the major anti-apoptotic proteins is sufficient to induce cell death even without involvement from major pro-apoptotic proteins. Importantly, our study also indicates that DRP-1 inhibition is a promising adjuvant for BH3 mimetics in melanoma treatment.