Loss of secreted frizzled-related protein-1 expression in renal cell carcinoma reveals a critical tumor suppressor function.

In this issue of Clinical Cancer Research , Gumz et al. report on the loss of secreted frizzled-related protein-1 (sFRP-1) in clear cell renal cell carcinoma (RCC; ref. 1). sFRP-1 is a member of a family of proteins that bind Wnts and can inhibit Wnt signaling during embryonic development and in the adult homeostasis (2, 3). Inappropriate Wnt signaling is a well-established mechanism of oncogenesis in several malignancies, and in these contexts, sFRPs have been viewed as potential tumor suppressors. Consistent with this hypothesis, the chromosomal loci of sFRPs have been associated with loss of heterozygosity in various tumor types (4, 5). Moreover, several studies have documented the loss of SFRP expression in cancer due to promoter hypermethylation (3). Recent literature, including the article by Gumz et al. (1), indicates that silencing of SFRP expression, particularly SFRP1 , is very common in RCC (6, 7). Importantly, Gumz et al. show that restoration of SFRP1 expression not only reduced the expression of Wnt target genes, but also markedly inhibited tumor cell growth in culture, soft agar, and xenografts in athymic nude mice (1). These findings strongly suggest that loss of sFRP-1 has a pivotal role in RCC, and that restoration of its expression could be of therapeutic benefit. Furthermore, analysis of SFRP gene hypermethylation may have diagnostic and prognostic value in the detection and management of RCC (6), which now claims more than 13,000 lives each year in the U.S. alone (8). Wnts have many diverse functions during embryogenesis, including a critical role in kidney development (9), and they contribute to stem cell renewal in the adult (2). Constitutive activation of the canonical Wnt/h-catenin pathway is common in many cancers, in which it is thought to be an early event in tumorigenesis, particularly in colorectal cancer (2). Stimulation of this pathway inhibits the degradation of cytosolic h-catenin, facilitating its accumulation in the nucleus where it associates with members of the T cell factor/lymphoid enhancer factor (TCF/LEF) family to function as a transcriptional activator (Fig. 1). Although activation of the pathway in cancer is usually attributed to deregulation of downstream effectors (e.g., h-catenin) or suppressors [e.g., adenomatous polyposis coli (APC) protein or Axin], autocrine mechanisms also have been described (10). Maximal stimulation probably involves a combination of both mechanisms (11). Other Wnt signaling cascades have been well-characterized, including a calcium/ protein kinase C mechanism implicated in metastatic melanoma (12) as well as a Rac/Rho-associated process that contributes to cell polarity and motility (13). All of these signaling events are mediated by Wnt receptors in the Frizzled (Fzd) family of seven-pass transmembrane proteins. In addition, the lipoprotein receptor– related proteins 5 and 6 (LRP5/6) function as Wnt coreceptors, specifically in the h-catenin pathway (Fig. 1). sFRPs possess a conserved Fzd-type cysteine-rich domain that binds Wnts and typically they antagonize Wnt signaling, presumably by preventing Wnt/Fzd interactions. Four of the five SFRP genes contain dense CpG islands (14), and a growing literature indicates that these genes often are silenced in cancer, primarily by promoter hypermethylation (3). The incidence of gene silencing is particularly high for sFRP-1, both with regard to the diversity of tumor types and the frequency within an individual type (3). Loss of sFRP-1 expression in breast cancer has been associated with decreased survival (15), and restoration of expression in colorectal tumor cell lines resulted in an attenuated tumor phenotype, manifested as diminished anchorage-independent growth in soft agar and increased apoptosis (11). Of note, this inhibition was observed in cells with mutations in APC or h-catenin, providing support for the idea that Wnt stimulation is needed to drive oncogenic signaling by these mutations, and that their full impact may result from the simultaneous activation of h-catenin and noncanonical Wnt pathways. These reports highlighted the potential importance of sFRP silencing in oncogenesis, and provided a strong impetus for the definitive functional experiments done by Gumz et al. Using a combination of genomic profiling, quantitative PCR, and immunohistochemistry, Gumz et al. documented a remarkably high incidence of sFRP-1 loss in clear cell RCC, with decreases also seen in papillary RCC but not chromophobe or oncocytoma (1). Thirteen Wnt-responsive genes were found to be dramatically up-regulated in RCC specimens. Reintroduction of sFRP-1 into clear cell RCC cell lines via stable cDNA transfection decreased the expression of these genes by 2to 3-fold. More impressive was the nearly complete inhibition of RCC cell growth in monolayer culture and soft agar assays. Additional experiments indicated that sFRP-1 expression decreased cell proliferation, but did not stimulate apoptosis, in contrast to the proapoptotic effect observed in colorectal tumor cell lines (11). Finally, Gumz and colleagues The Biology Behind