Secreting Tumor Suppression

Abstract

Cellular senescence limits the proliferative capacity of damaged cells and thereby acts as an intrinsic mechanism of tumor suppression. In this issue, Wajapeyee et al. (2008) identify insulin growth factor binding protein 7 (IGFBP7) as a secreted factor that mediates senescence induced by oncogenic BRAF in normal melanocytes. In addition, IGFBP7 triggers apoptosis in cells that have progressed to melanoma, suggesting a new approach for melanoma treatment.

Replicative senescence was first described as a permanent state of proliferative arrest occurring in cells after extended culture in vitro (Hayflick, 1965). Whereas replicative senescence is triggered by telomere erosion, several other stress-inducing factors initiate a similar process, which occurs more rapidly than replicative senescence and without extensive cell division. This process, generally referred to as “cellular senescence,” acts as a program to limit the proliferative capacity of damaged cells. Stimuli that induce cellular senescence include DNA damage, oxidative stress, chemotherapeutic drugs, and expression of certain activated oncogenes. Wajapeyee et al. (2008) now report that a secreted factor, insulin growth factor binding protein 7 (IGFBP7), induces cellular senescence in melanocytes that contain activating mutations in the BRAF oncogene.

The first oncogene shown to trigger senescence was a tumor-derived allele of H-RAS (Serrano et al., 1997). At that time, the transforming activity of RAS in immortalized rodent cells was well established, and its ability to induce senescence in primary cells explained why these cells could not be transformed by RAS alone but required additional “immortalizing” factors, such as loss of tumor suppressor genes. Subsequent studies revealed that this occurs via signaling through the MAPK cascade, and thus activated forms of RAF and MEK also produce similar phenotypes. More recent reports suggest that RAS-induced senescence involves a DNA-damage response induced by replication stress (Campisi and d’Adda di Fagagna, 2007). Thus, senescence may act to counter the tumor-promoting effects of hyper-proliferative mutations and consequently is a “built-in” or intrinsic mechanism of tumor suppression (Lowe et al., 2004). Consistent with this view, execution of RAS-induced senescence requires the p53 and Rb tumor suppressor pathways (Serrano et al., 1997).

Although the physiological relevance of oncogene-induced senescence has been debated, recent reports indicate that this process acts as a potent barrier against tumorigenesis (Narita and Lowe, 2005). As one example, melanocytic nevi (moles) are premalignant lesions that are extremely stable and rarely progress, despite consisting of melanocytes containing activating mutations in the BRAF gene (predominantly V600E, a glutamic acid to valine substitution at position 600). Indeed, nevi contain cells showing hallmarks of senescence, and expression of BRAFV600E in cultured fibroblasts or melanocytes induces senescence. Melanomas appear to acquire alterations that enable them to evade senescence. Consequently, the senescence response halts the growth of benign neoplasms, thereby limiting their malignant progression.

In this issue, Wajapeyee et al. identify one mechanism by which oncogenic BRAF triggers cellular senescence in melanocytes. They conducted a genome-wide RNA interference (RNAi) screen to identify genes that are required for BRAFV600E to inhibit proliferation of human diploid fibroblasts and primary melanocytes. This screen led to the identification of a secreted protein, IGFBP7, that is required for the process. Expression of BRAFV600E in melanocytes induces synthesis and secretion of IGFBP7, which then triggers senescence as an autocrine/paracrine factor. Intriguingly, melanoma cells harboring activated BRAF genes often silence IGFBP7 expression, suggesting one way in which they bypass oncogene-induced senescence.

How does IGFBP7 execute its function? Wajapeyee et al. suggest that IGFBP7 acts through a negative feedback loop that attenuates the BRAF-MEK-MAPK signaling cascade, leading to senescence in primary melanocytes and apoptosis in melanoma cells containing BRAF mutations. A conceptually similar but mechanistically different feedback was recently suggested to account for senescence induced by disruption of the neurofibromatosis tumor suppressor, which also deregulates MAPK signaling (Courtois-Cox et al., 2006). By contrast, IGFBP7 induces apoptosis in melanoma cells with mutations in BRAF through increasing the expression of proapoptotic proteins, SMARCB1 and BNIP3L. Why IGFBP7 induces senescence in primary melanocytes and apoptosis in melanomas remains a puzzle; nevertheless, studies suggest that oncogene-expressing cells that have bypassed senescence, such as cancer cells, are in general more susceptible to induction of apoptosis (Lowe et al., 2004).

Previous work demonstrates that senescent cells display dramatic changes in their transcriptional profiles that distinguish them from growing or quiescent cells. In many instances, these changes involve upregulation of growth inhibitory proteins and downregulation of cell-cycle-promoting genes. However, in addition to these intuitive effects, senescent cells also upregulate a class of genes encoding secreted factors such as matrix metalloproteinases, protease modulators, growth factors, and chemokines. Although formerly used merely as a marker of senescence, the “senescence-associated secretory phenotype” may be biologically significant, particularly with regards to the role of senescence in vivo. For example, secreted factors from senescent fibroblasts promote growth of adjacent epithelial cells (Krtolica et al., 2001), suggesting that the secreted proteins produce a microenvironment that favors malignant progression. By contrast, the secretory phenotypes of senescent tumor cells can target these cells for clearance in vivo through an innate immune response (Xue et al., 2007), suggesting a mode of immune surveillance that can have an antitumor effect.

Yet, until recently, the role of proteins secreted from senescent cells was not linked to the antiproliferative program per se. However, a recent study indicates that one secreted protein, plasminogen activator inhibitor-1 (PAI-1), is a p53 target gene and acts together with p21, a cyclin-dependent kinase inhibitor, to trigger senescence in both murine and human fibroblasts (Kortlever et al., 2006). The findings by Wajapeyee et al. are even more remarkable because they suggest that one of these secreted factors (IGFBP7) is both necessary and sufficient to establish a senescent state. In both studies, the secreted factors ultimately interfered with mitogenic signaling, although in the case of PAI-1 this impacted the PI3-kinase pathway, whereas with IGFBP7 the effects were largely on MAPK signaling.

The differences between the studies noted above suggest that the signaling networks that trigger senescence can be highly context dependent, and it remains to be determined how universal the IGFBP7 mechanism is in the control of senescence and its relationship to other more established pathways of senescence. Moreover, it is not known whether other senescence-promoting lesions, for example, activation of RAS or loss of PTEN, also rely on IGFBP7. In fact, despite the canonical view that RAS and RAF function in a linear pathway, IGFBP7 has no effect on melanomas with activated RAS mutations. Furthermore, the two major senescence regulators, p53 and p16^INK4a, are apparently not involved in BRAF-induced senescence. These observations suggest key differences between RAS and BRAF signaling during senescence and perhaps other biological processes as well.

The new study has important implications for our understanding of melanoma progression (Figure 1). Indeed, the authors show that increases in IGFBP7 occur in human benign nevi containing BRAF mutations, whereas this is not observed in human melanomas. Thus, IGFBP7 expression appears to be invariably lost during melanoma progression, presumably allowing melanocytes to evade the senescence response to BRAF signaling. Undoubtedly, mutations that activate or repress cellular senescence will be crucial in the progression of many other human malignancies.

Figure 1. IGFBP7 and BRAF-Induced Cellular Senescence.

Early during melanoma progression, melanocytes acquire BRAF mutations that stimulate proliferation but also trigger senescence, in part, via enhanced secretion of insulin growth factor binding protein 7 (IGFBP7). Silencing of IGFBP7 enables BRAF-expressing melanocytes to escape senescence and progress toward melanoma. The resulting melanoma cells remain sensitive to the induction of apoptosis by addition of IGFBP7.

A great promise of research into the underlying mechanisms of apoptosis and cellular senescence is that this will suggest strategies to harness the power of intrinsic tumor-suppressive mechanisms for improved cancer therapies. Clearly, Wajapeyee et al. suggest that this may be possible for melanoma, an extremely aggressive and chemo-resistant cancer. Specifically, the authors show that IGFBP7 has a potent antitumor effect on xenograft tumors derived from BRAF mutated melanoma cells but has little if any effect on tumors containing wild-type BRAF. Perhaps this suggests that cells with BRAF mutations become addicted to loss of IGFBP7, much as has been implied for certain tumor cells expressing the BCR-ABL or EGFR oncogenes (Sharma and Settleman, 2007). Whether such stunning antitumor effects will be observed in human patients remains to be determined; nevertheless, these results demonstrate that key tumor-suppressive mechanisms can operate via non-cell-autonomous mechanisms and provide proof-of-principle that this can be exploited therapeutically. The results also emphasize the utility of nonbiased genetic screening as an entry point to unexpected, and clinically useful, realms of biology.