Individualized strategies to target specific mechanisms of disease in malignant melanoma patients displaying unique mutational signatures

Abstract

Targeted treatment of advanced melanoma could benefit from the precise molecular characterization of melanoma samples. Using a melanoma-specific selection of 217 genes, we performed targeted deep sequencing of a series of biopsies, from advanced melanoma cases, with a Breslow index of ≥ 4 mm, and/or with a loco-regional infiltration in lymph nodes or presenting distant metastasis, as well of a collection of human cell lines. This approach detected 3-4 mutations per case, constituting unique mutational signatures associated with specific inhibitor sensitivity. Functionally, case-specific combinations of inhibitors that simultaneously targeted MAPK-dependent and MAPK-independent mechanisms were most effective at inhibiting melanoma growth, against each specific mutational background. These observations were challenged by characterizing a freshly resected biopsy from a metastatic lesion located in the skin and soft tissue and by testing its associated therapy ex vivo and in vivo using melanocytes and patient-derived xenografted mice, respectively. The results show that upon mutational characterization of advanced melanoma patients, specific mutational profiles can be used for selecting drugs that simultaneously target several deregulated genes/pathways involved in tumor generation or progression.

Keywords: BRAF; MAPK; melanoma; somatic mutations; targeted therapy.

Figure 1. In silico-targeted mutational profiling of advanced melanoma patients

A. Meta-analysis showing the average number of mutated genes per case with the potential to guide targeted therapy. Original mutational data from cell lines (red bar) were obtained from the Cancer Cell Line Encyclopedia website (see Material and Methods); mutational data from patients (grey bars) were obtained from Nilolaev [19], Hodis [18], Stark [14] and Berger [13]; Black bar, shows the average frequency of mutations amongst all data sets. B. Percentage of hits in A) involved in the indicated signaling pathway.

Figure 2. Effects of specific targeted therapy guided by mutational signature

A. Proliferation analysis of A375 cells at 0, 24 and 48 h. Cells were seeded in 96-well plates and treated with the indicated concentrations of each inhibitor: B-RAFi (V: Vemurafenib), FGFR2i (Va: Vargatef), and mTORi (E: Everolimus). B. Western blots using whole cell lysates from starved A375 cells incubated for 1 h with control vehicle (DMSO) or the indicated concentration of each inhibitor. The figure shows a representative experiment using P-ERK1/2, ERK1/2, P-p38, p38, P-S6 and S6 antibodies, as indicated. C. Proliferation analysis of A375 cells in the same conditions as in A), but incubated with control vehicle (DMSO) or the IC₅₀ concentration of the indicated inhibitor alone (blue lines), or in a double (green lines) or triple combination (red line). N = 6. Error bars show the SEM. D. DNA synthesis using Click-iT^® EdU in exponentially growing A375 cells seeded in an 8-well glass and incubated for 48 h with control vehicle (DMSO) or the indicated inhibitor or combination of inhibitors, as in C). Graph bars show percentage of low (clear red) or high (intense red) EdU-stained cells in three photographic fields from a representative experiment. E. Representative pictures of each treatment condition showing the nucleus of the total number of cells (blue dots) and EdU-positive cells (red dots). F and G. Western blots of whole cell lysates of the indicated cells. Cells were starved overnight and incubated for 1 h with control vehicle (DMSO), or the indicated inhibitor, or a combination of inhibitors under the same conditions as in C). Figures show representative experiments using P-ERK1/2, ERK1/2, P-p38, p38, P-S6 and S6 antibodies, as indicated. Bar graphs show the values of three independent experiments in G). Error bars indicate the SEM.

Figure 3. Increased effects of targeted therapy against an appropriate mutational background

A. Proliferation analysis of exponentially growing A375 (BRAF+), SKMEL2 (NRAS+), HT144 (BRAF+), MALME (BRAF+), MEWO and MELANOMA17 (BRAF+) cells treated with vehicle (DMSO) or the IC₅₀ concentration (calculated for A375 cells) of the indicated inhibitor alone, in a double or triple combination for 48 h. N = 6. Error bars show the SEM. B. Western blots of whole cell lysates of the indicated cells. Cells were starved overnight and incubated for 1 h with control vehicle (DMSO), or the indicated inhibitor, or a combination of inhibitors under the same conditions as in A). Figure shows a representative experiment.

Figure 4. In vivo effects of a targeted therapy combining MAPK-ERK-dependent and MAPK-ERK-independent mechanisms of inhibition

A. Xenografted tumor growth-derived A375 cells injected subcutaneously in 48 BALB/C nude mice. Tumor size was monitored until a volume of 100 mm³ was obtained, whereupon mice were assigned to four treatment groups: 1) Control (DMSO, blue line); 2) BRAFi (V) (orange line); 3) FGFR2i (Va) + mTORi (E) (green line); and 4) BRAFi (V) + FGFRi (Va) + mTORi (E) (red line). Mice were treated daily as indicated (see Materials and Methods for further details) and tumor volumes were measured until day 13, at which point the experiment was ended. Data were obtained from the 12 control, 11 (V), 7 (Va+E) and 10 (V+Va+E) mice that survived the entire process. Error bars indicate the SEM. B. Representative pictures illustrating the effects of the indicated treatment on the xenografted tumors that had been resected or were still in the mice. C. H&E staining of representative tumor sections from five representative mice for each treatment condition. Tumor sections were analyzed for Ki67-positive staining D. or by the number of mitoses E. Data are averages of five section cuts in each mouse. Error bars indicate the SEM. F. Immunohistochemical (IHC) analysis in tumors corresponding to the indicated treatment, as in C), using an anti-phospho ERK antibody stain. G. Tumor sections were analyzed for phosphor-ERK-positive staining. Results are the averages of five section cuts per mouse. Error bars indicate the SEM.

Figure 5. A pre-clinical example of targeted therapy guided by a specific mutational signature in melanoma patient 17

A. Schematic representation of the work performed with a freshly resected biopsy from patient 17. B. Sanger sequencing of BRAF (above) and MAPK7 (ERK5; below) oncogenes in genomic DNA from control cells or isolated melanocytes from patient 17 (MELANOMA17 cells). C. Western blots of whole-cell lysates from starved MELANOMA17 cells incubated for 1 h with control vehicle (DMSO) or the indicated concentration of each inhibitor BRAFi (V: Vemurafenib) or ERK5i (X: XMD-8–92). The figure shows a representative experiment using P-ERK1/2, ERK1/2, ERK5 and tubulin antibodies, as indicated. D. Proliferation analysis of MELANOMA17 cells at 0, 24 and 48 h. 3 × 10³ cells/well were seeded in 96-well plates and treated with control (DMSO) (black line), or an IC₅₀ concentration of B-RAFi (V) (blue line) or ERK5i (X) (red line), alone or in combination (green line). N = 6. Error bars show SEM. E. Proliferation analysis of MELANOMA17, A375, MEL JUSO, SKMEL2, and MEWO cells at 0, 24 and 48 h, under the same conditions as in D). N = 6. Error bars show the SEM. F. Western blots using whole cell lysates of the indicated cells. Cells were starved overnight and incubated for 1 h with control vehicle (DMSO), or the indicated inhibitor, or a combination of inhibitors using the same concentrations as in E). Representative experiment using anti-P-ERK1/2 and anti-ERK1/2 antibodies. G. Tumor growth derived from 2-mm³ MELANOMA17-derived tumor fragments implanted subcutaneously in 30 NSG mice (Jackson Laboratories). Tumors were monitored until they attained a volume of 100 mm³, whereupon mice were assigned to four comparable treatment groups: 1: Control (DMSO, black line), 2: BRAFi (V) (blue line), 3: ERK5i (X) (red line) and 4: BRAFi (V) + ERK5i (X) (green line). Mice were treated daily as indicated (see Materials and Methods for further details) and tumor volumes were monitored until day 13, at which point the experiment was ended. Data were obtained from five survivor mice from each treatment group. Error bars indicate the SEM. The figure shows a representative image from treated tumors that were still in the mice (above) or had been freshly resected (below). H. Bar graph of average changes in tumor volume (left) and mass (right) for each treatment condition. N = 5. Error bars indicate the SEM. I. Examples of IHC analysis of tumors corresponding to the treatment indicated in D) using anti-phospho ERK antibody staining. Bar graphs show results for tumor sections analyzed for phosphor-ERK-positive staining. Data are the averages of five section cuts per mouse. Error bars indicate the SEM.