The FAT epidemic: a gene family frequently mutated across multiple human cancer types.

In recent years, mutational surveys published by our group, The Cancer Genome Atlas, and others have revealed highly prevalent mutations and deletions targeting an intriguing family of genes encoding transmembrane proteins: the FAT family (FAT1, FAT2, FAT3 and FAT4). In particular, FAT1 and FAT4 appear to be the most frequently altered. Both genes harbor recurrent mutations and deletions in multiple types of human cancer, including glioblastoma (FAT1, 20%), colorectal (FAT1, 9%; FAT4, 20%), head and neck squamous cell (FAT1, 12%; FAT4, 10%), gastric (FAT4, 20%), serous ovarian (FAT1, 7%; FAT4, 3%) and pancreatic (FAT1, 4%; FAT4, 8%) cancers.-

These somatic alterations are intriguing, given the closely related Drosophila Fat gene, which has been identified as a tumor suppressor gene. In the fly, Fat acts upstream of the Hippo signaling pathway, which restrains cell growth; loss of Fat causes overgrowth of larval imaginal discs in flies. Fat is also a key developmental gene, controlling cell polarity along the apical-basal axis. The human FAT genes are all closely related, although recent phylogenetic analyses reveal that the closest human ortholog of Drosophila Fat is in fact FAT4, while FAT1, FAT2 and FAT3 are actually most closely related to the Drosophila gene Fatl (Fat-like). Nevertheless, all four human genes bear tight similarity to the Drosophila Fat.

The FAT genes all encode large membrane proteins called protocadherins, which have around 30+ cadherin repeats, four or five EGF-like domains and laminin G-like domains. The FAT proteins are present in a wide range of species and are highly conserved. They are believed to represent the first form of cadherins and were accordingly named “protocadherins.” In certain cell types, FAT proteins can be concentrated at lamellipodia, filopodia and sites of cell-cell contact. Like the classical cadherins, the protocadherins often serve dual roles in cell adhesion and signaling.

FAT4 is most closely related to Fat and mediates similar key developmental functions, such as planar cell polarity. It appears to have growth- and invasion-suppressive properties in several cancer cell lines, but the mechanisms remain obscure. In Drosophila, Fat is the apical member of the Hippo signaling pathway, which regulates organ growth and cell cycle progression in response to cell density. Although definitive data are still pending in the context of cancer, FAT4 appears to participate in mammalian Salvador/Warts/Hippo signaling during development. It is therefore quite possible that the frequent FAT4 alterations observed in cancer lead to pro-proliferation signals through a loss of proper regulation of the Hippo pathway.

For FAT1, more complete genomic, functional and mechanistic evidence has recently emerged, revealing the molecular basis of its tumor-suppressive function. This gene is a target within a highly prevalent region of deletion at 4q35 observed across many types of human cancer. We have identified mutations in glioblastoma, colorectal and head and neck samples at rates of 7–20%. FAT1 appears to be among the most frequently mutated genes in squamous cell carcinoma of the head and neck. Both in vitro and in vivo, depletion of FAT1 leads to markedly accelerated cell growth and proliferation, while expression of FAT1 robustly suppresses tumor growth. These growth-suppressive effects are abrogated when mutations observed in tumors are present. Just as classical cadherin proteins can bind to β-catenin and regulate its transcriptional activity, FAT1 also binds β-catenin and limits its translocation to the cell nucleus. Mutations in FAT1’s intracytoplasmic domain result in a loss of this ability to regulate β-catenin. Therefore, loss of FAT1 in cells activates the Wnt signaling pathway, unleashing β-catenin-dependent transcriptional activity and upregulating pro-growth wnt transcriptional targets. Consistent with this, primary cancer samples with FAT1 alterations are defined by significant enhancement of Wnt signaling. The growth-suppressive functions of FAT1 are mediated by its intracytoplasmic, β-catenin binding domain, but the extracellular domain also mediates cell-cell adhesion, which may be a secondary mechanism by which FAT1 loss promotes tumor growth.

The Wnt/β-catenin pathway has been causally linked to multiple types of cancer. For example, in colorectal cancer, the overwhelming majority of tumors are defined by alteration of the core genes, APC, CTNNB1 and AXIN1/2. However, in non-colorectal tumors, these genes are infrequently or never mutated. The precise causes of Wnt pathway activation in these types of cancer have not been elucidated. New data now implicates FAT1 mutation as a driver of Wnt activation in many of these tumors, and point to FAT1 as a potential molecular determinant for guiding use of new small-molecule inhibitors of Wnt signaling.

Interestingly, FAT1 appears to play multiple, seemingly opposing, roles in development and cell growth. While the protein has a strong tumor-suppressive effect, it also binds to Ena/VASP, thereby promoting actin polymerization and cell motility. Indeed, in experimental systems, FAT1 promotes lamellipodial dynamics and cell migration and invasion. These different functions may be vestiges of FAT1’s role in development, where it may play a directional role in guiding organ development. For these reasons, it is conceivable that the function of FAT1 in tumorigenesis is multifaceted, such that the gene may not operate as a tumor suppressor via the same mechanisms in all cellular contexts.

FAT2 and FAT3 have not been well characterized to date, although each gene has been observed to be undergo mutation in approximately 10% of colorectal and lung squamous cell cancers., These two genes have more similarity to FAT1 than FAT4, but it remains to be seen whether these protocadherins are also able to modulate Wnt signaling. As additional mutational data in diverse human cancers is reported, the precise implications of the recurrent alterations targeting these intriguing, large, ancient proteins will be more thoroughly illuminated. (Fig. 1)

Figure 1. Model of FAT1 function. When FAT1 is present, β-catenin is held at the cell membrane. When FAT1 is inactivated by mutation or deleted in cancers, an excess of β-catenin is present in the cytoplasm. Some β-catenin is then able to enter the nucleus and cooperate with TCF transcription factors to activate gene expression of Wnt target genes. The APC/GSK-3/Axin complex can potentially target β-catenin for degradation. However, excess β-catenin caused by FAT1 inactivation may not all be degraded by the APC/GSK-3/Axin complex or the complex may not be as active in non-colorectal cells.