Lack of correlation between IGFBP7 expression and BRAF mutational status in melanoma.

TO THE EDITOR Although the incidence of melanoma is still rising in many countries, the exact molecular pathogenesis for this tumor is unknown. Consequently, the publication of the study by Davies et al. (2002) reporting a high frequency of activating BRAF mutations in melanoma attracted much attention in the research community. Indeed, many independent groups have confirmed the high frequency of BRAF mutations in patients with melanoma; the mutation data on the COSMIC website (http://www.sanger.ac.uk/genetics/CGP/cosmic/) indicate that BRAF mutations were detected in 46% of melanoma patients (n=3,634), with BRAF V600E being the predominant mutation, found in 37.2% of the analyzed patients. However, given that in benign melanocytic nevi the frequency of BRAF mutations is similarly high (35% total mutations; 29.5% BRAF V600E; n=830, including 358 Spitz and blue nevi having a BRAF mutation frequency of ~7%), BRAF mutations alone are not sufficient to initiate malignant melanoma. This thesis is sustained by experimental data showing that ectopic expression of constitutive active BRAF leads to senescence of primary cells (Michaloglou et al., 2005). An autocrine/paracrine regulatory loop explaining this observation was recently described by Wajapeyee et al. (2008), who demonstrated the presence in normal cells of a basal expression of insulin-like growth factor–binding protein 7 (IGFBP7) that inhibits BRAF–MEK–ERK signaling and thereby restrains apoptosis. In nevi, BRAF mutations activating the MAP kinase pathway lead to an increased expression of IGFBP7, which not only inhibits BRAF–MEK–ERK signaling but also activates senescence. In contrast, in BRAF-mutated melanoma lesions, IGFBP7 expression is absent, enabling the cells to escape from senescence and thereby exhibit uncontrolled proliferation. Besides delivering the explanation for the observed effects of constitutive BRAF activation, the article by Wajapeyee and colleagues demonstrates a potential therapeutic application of this knowledge: treatment with recombinant IGFBP7 of BRAF-mutated melanoma cells triggered their apoptosis both in vitro and in vivo in a xenotransplantation model. Similarly, reactivation of IGFBP7 by DNA demethylation inhibits colon cancer cell growth in vitro (Lin et al., 2008). With this therapeutic implication in mind, we further scrutinized the expression of IGFBP7 in melanoma. To this end, we stained formalin-fixed tissues obtained from 41 primary tumors and 16 metastases for IGFBP7 expression. Unlike Wajapeyee and colleagues, however, we did not detect a clear demarcation of IGFBP7 expression between BRAF wt (wild type) and mutated melanoma lesions (Figure 1). Indeed, although the frequency of IGFBP7-expressing cells is lower in BRAF-mutated melanoma lesions overall, the whole diversity of IGFBP7 expression from absent to present in 100% of tumor cells was observed in both wt and BRAF-mutated lesions (P=0.1609; Mann–Whitney test). Accordingly, western blot analysis of short- and long-term cultured melanoma cell lines did not show any correlation between IGFBP7 protein expression and BRAF status either (Figure 2). Heterogeneous IGFBP7 expression could still be in accordance with the reported stringent IGFBP7/BRAF correlation, assuming a corresponding heterogeneity in the BRAF status. However, the homogeneous IGFBP7 expression in some of the BRAF-mutated tumors, as well as the lack of correlation for the cell lines, with one of them being hemizygote for the V600E BRAF mutation, argues against an obligatory downregulation of IGFBP7 expression in BRAF-mutated melanoma cells. Rather, it seems that loss of IGFBP7 expression is not the only way to overcome MAPK pathway–induced senescence. In this regard, dysregulation of the other 16 candidates detected by Wajapeyee et al. (2008), including BNIP3L, FOXA, and NF2, may be alternative mediators for overcoming senescence. Thus, detailed studies scrutinizing different modi operandi by which melanoma cells overcome BRAF-induced senescence are needed before the potential of IGFBP7 substitution for treatment of melanoma can be estimated. The authors state no conflict of interest. We thank Claudia Siedel for excellent technical assistance. This work was supported by the Deutschen Krebshilfe (Verbundprojekt Melanom Teilprojekt 11).