Background

In melanoma, dysregulation of the MAPK pathway, usually via BRAF(V600) or NRAS(Q61) somatic mutations, leads to constitutive ERK signaling. While BRAF inhibitors are initially effective for BRAF-mutant melanoma, no FDA-approved targeted therapies exist for BRAF-inhibitor-resistant BRAF(V600), NRAS mutant, or wild-type melanoma.

Methods

The 50% inhibitory concentration (IC50) of SCH772984, a novel inhibitor of ERK1/2, was determined in a panel of 50 melanoma cell lines. Effects on MAPK and AKT signaling by western blotting and cell cycle by flow cytometry were determined.

Results

Sensitivity fell into three groups: sensitive, 50% inhibitory concentration (IC50) < 1 μM; intermediately sensitive, IC50 1-2 μM; and resistant, >2 μM. Fifteen of 21 (71%) BRAF mutants, including 4 with innate vemurafenib resistance, were sensitive to SCH772984. All three (100%) BRAF/NRAS double mutants, 11 of 14 (78%) NRAS mutants and 5 of 7 (71%) wild-type melanomas were sensitive. Among BRAF(V600) mutants with in vitro acquired resistance to vemurafenib, those with MAPK pathway reactivation as the mechanism of resistance were sensitive to SCH772984. SCH772984 caused G1 arrest and induced apoptosis.

Conclusions

Combining vemurafenib and SCH722984 in BRAF mutant melanoma was synergistic in a majority of cell lines and significantly delayed the onset of acquired resistance in long term in vitro assays. Therefore, SCH772984 may be clinically applicable as a treatment for non-BRAF mutant melanoma or in BRAF-mutant melanoma with innate or acquired resistance, alone or in combination with BRAF inhibitors.

Histone deacetylase inhibitors (HDACi) have been reported to increase tumor antigen expression, and have been successfully tested as adjuvants for melanoma immunotherapy in mouse models. In this work, we tested the effects of a pan-HDACi on human lymphocytes and melanoma cell lines. Effects of the pan-HDACi panobinostat (LBH589) on cell viability, cell cycle, apoptosis, and DNA damage were determined in peripheral blood mononuclear cells (PBMC) from 2 healthy donors, 13 patients with metastatic melanoma, 2 bone marrow samples from patients with different malignances, and 12 human melanoma cell lines. Intracellular signaling in lymphocytes, with or without cytokine stimulation, was analyzed by phospho-flow cytometry in one of each type. The IC50 in PBMCs was <20 nmol/L compared with >600 nmol/L in melanoma cell lines; >40% apoptotic cell death in PBMCs versus <10% in melanoma cell lines was seen at the same concentration. Phospho-histone variant H2A.X (pH2A.X) increased 2-fold in healthy donor PBMCs at 1 nmol/L, whereas the same effect in the melanoma cell line M229 required 10 nmol/L. pH2A.X was inhibited slightly in the PBMCs of 3 patients with metastatic melanoma at 1 nmol/L and in the melanoma cell line M370 at 10 nmol/L. Panobinostat inhibited phospho-STAT1/3/5/6, -p38, -ERK, -p53, -cyclin D3, and -histone H3 in flow cytometry-gated healthy donor B and T cells, whereas it induced up to 6-fold activation in patients with metastatic melanoma and bone marrow samples. In human lymphocytes, panobinostat alters key lymphocyte activation signaling pathways and is cytotoxic at concentrations much lower than those required for melanoma antitumor activity, resulting in an adverse therapeutic window.