C-Raf is associated with disease progression and cell proliferation in a subset of melanomas

Abstract

Purpose: Raf-kinases include three major isoforms. Although the role of B-Raf in melanoma is well established, little is known about C-Raf. We studied effects of C-Raf knockdown in vitro and assessed expression of C-Raf in a large cohort of melanomas and nevi.

Experimental design: Using specific siRNAs, we knocked down C-Raf expression, and determined the effect on viability, MAP extracellular signal-regulated kinase (ERK)/ERK kinase signaling, and apoptosis in seven melanoma cell lines. We determined the IC(50) of the C-Raf inhibitors sorafenib and GW5074, and studied the effects of GW5074 on cell signaling. Using an automated method to measure in situ protein expression, we quantified C-Raf expression in 263 nevi and 523 melanomas.

Results: C-Raf was knocked down in three cell lines with detectable phospho-C-Raf, resulting in decreased viability in two of the three (YULAC and YUROB). This resulted in decreased Bcl-2 expression and phospho-Bad cleavage, without affecting phospho-MEK and phospho-ERK. Sensitivity to sorafenib and GW5074 varied. GW5074 inhibited mitogen-activated protein kinase signaling without Bcl-2 and phospho-Bad down-regulation. C-Raf was highly expressed in melanomas compared with nevi (P < 0.0001), and no nevi had high C-Raf expression. C-Raf expression was higher in metastatic than primary specimens (P = 0.0225).

Conclusions: C-Raf siRNA knock-down results in decreased viability of YULAC (B-Raf(V600K)) and YUROB (B-Raf(WT)) melanoma cells, likely mediated by Bcl-2 inhibition rather than mitogen-activated protein kinase inhibition. Cotargeting C-Raf and parallel pathways might be an effective therapeutic approach for melanoma. C-Raf expression is up-regulated in a subset of melanomas but not in nevi, suggesting that it might be a valuable diagnostic marker and therapeutic target.

Figure 1. C-Raf silencing in melanoma cell lines

YUSAC, YUMAC, YULAC, YUGEN, YUFIC, YUROB and MEL501 cells were transfected with one of two siRNA clones directed against C-Raf (lanes 3 and 4 respectively) or a non-targeting control siRNA (TS1058), (lanes 5). Following treatment, cells were lysed and proteins were extracted, electrophoresed and probed for expression of C-Raf. Equal protein gel loading was confirmed by probing for expression of β-Actin.

Figure 2. Effect of C-Raf silencing on cell viability

YUSAC, YUMAC, YULAC, YUGEN, YUFIC, YUROB and MEL501 cells were transfected with 75 nM of either C-Raf siRNA or a control siRNA (TS1058) for 96 hours (A), 72 hours (B) and 48 hours (C). Cell viability was determined by the CellTiter 96^® Aqueous One Solution Cell Proliferation Assay. Following 96 hours treatment, 75nM C-Raf siRNA inhibited cell viability of YULAC and YUROB cell lines by 43-58% and 50-65% respectively. At 72 hours of incubation C-Raf siRNA inhibited cell proliferation of the YULAC cell line by 23-39%. Viability of the other cell lines was less affected at all time points. 75nM of the control siRNA did not affect C-Raf expression or viability.

Figure 3. Effect of C-Raf silencing on protein expression

YUSAC, YUMAC, YULAC, YUGEN, YUFIC, YUROB and MEL501 cells were transfected with one of the C-Raf siRNA clones at 75nM. Following 96 hour treatment, cells were lysed and proteins were extracted, electrophoresed and probed for expression of B-Raf, C-Raf, phospho-MEK1/2 (Ser 221), phospho-ERK1/2 (Thr202/Tyr204), Bcl-2, phospho-Bad (Ser112) following stripping of the blots. Equal protein gel loading was confirmed by probing for expression of β-Actin. C-Raf knock down resulted in decreased levels of Bcl-2 and phospho-Bad in YULAC and YUROB cells, while it had less or no effect at all in other cell lines.

Figure 4. Membranous and cytoplasmic C-Raf expression in melanoma histospots

(A) AQUA uses a cocktail of S100 and HMB45 to define the tumor mask (bottom-left), 4,6-diamidine-2-phenylindole to define the nuclear compartment (bottom-right) and Cy5 (top-right) for the target (C-Raf). The cytoplasmic compartment is generated by subtracting the nuclear compartment from the S100/HMB45 mask (top-left). C-Raf expression is then measured within the nuclear and cytoplasmic compartments within the tumor mask, and each spot is assigned a score based on pixel intensity per unit area. C-Raf expression in this example corresponds to an AQUA score of 35.922 (measured in the cytoplasmic compartment) and is shown at 10X magnification (top-right panel A) and (B) 40X magnification. (C) Regression plots for scores from the two sets of melanoma arrays for C-Raf. Each array contains histospots from the same patients, taken from different areas of the tumor.

Figure 5. C-Raf expression in nevi and malignant specimens

(A) Comparison of C-Raf expression (Cytoplasmic AQUA scores) in nevi, primary and metastatic specimens. AQUA scores are represented as percentages of total nevi, primaries and metastases, respectively, by deciles. 99.57% of nevi have AQUA scores between 0 and 10 while only 0.43% had scores between 10 and 20. AQUA scores higher than 20 correspond only to either primary or metastatic specimens. (B) Comparison of C-Raf expression in nevi and all malignant specimens (metastatic and primary melanomas). Unpaired t tests showed that C-Raf expression was significantly higher in malignant versus benign tissue cores. (C) Comparison between high and low C-Raf expression in primary and metastatic melanoma specimens. We defined high C-Raf expression for malignant specimens as expression that was higher than the 95th percentile score for nevi. This score was 7.537.