To date, approximately 80 common fragile sites (CFS) have been identified in the human genome (Ruiz-Herrera et al. 2006). Unlike rare fragile sites, which map to sites that are genetically altered in individuals, CFS occur at consistent genomic locations in particular cell types, which are defined by the paucity of replication initiation events at these loci (Letessier et al. 2011). DNA breaks at CFS, referred to as expression of the fragile site, can be induced via treatment with inhibitors of DNA replication, carcinogens, and environmental stresses (Lukusa and Fryns 2008). Several CFS colocalize with common deletions and translocations in tumors. Early efforts to find tumor suppressor genes encoded within fragile sites focused on FRA3B at chromosomal location 3p14.2, the most frequently expressed human CFS. Perturbations at 3p14.2 had been noted in many cancers and the site includes the t(3;8) translocation breakpoint associated with hereditary clear cell renal carcinoma (cRCC) (Cohen et al. 1979; Dreijerink et al. 2001). Positional cloning identified the fragile histidine triad gene, FHIT, as the gene that spans FRA3B (Ohta et al. 1996). In addition to losses of FHIT on the short arm of chromosome 3, many cRCC malignancies have been subject to gene losses including RASSF1A at 3p21.3 (Dreijerink et al. 2001) and VHL at 3p25 (Latif et al. 1993).
FHIT Is a Tumor Suppressor Gene
Numerous cancer cell lines and primary tumors exhibit both homo- and hemizygous deletions within the FHIT gene. These deletions are concomitant with loss of mRNA expression in esophageal, stomach, colon, kidney, breast, and lung cancer cells. Immunohistochemistry of several types of primary tumors, including esophageal cancers, colon lesions, and cervical and lung cancers show frequent loss or absence of Fhit protein expression (Saldivar et al. 2010). This evidence notwithstanding, acceptance of FHIT as a tumor suppressor gene initially met with resistance, as loss of genetic material at a fragile site could be attributed as a consequence of genomic instability rather than a cause of cancer. However, accumulated evidence indicates that loss of Fhit is selected in carcinogenesis.
Mouse modeling and functional analysis has been used to show that FHIT mutations are drivers of malignancy. The murine Fhit locus shares key similarities with human FHIT: Fhit encompasses a fragile site, Fra14A2, and Fhit expression is altered in several murine cancer cell lines, suggesting that a mouse model of Fhit inactivation would be relevant to human disease (Glover et al. 1998). Mice heterozygous and homozygous at Fhit alleles have been established. These mice are fertile, live normal life spans, and do not present with gross defects, although an immune deficiency related to a low granulocyte count has been noted (Zanesi et al. 2001). Fhit heterozygous mice possess increased susceptibility to development of carcinogen-induced tumors. Oral administration of the carcinogen N-nitrosomethylbenzylamine (NMBA) produces tumors in 25% of a wild-type cohort of mice, while administration of an equivalent dose produces tumors in 100% of Fhit heterozygotes (Fong et al. 2000). Fhit heterozygous and Fhit-deficient mice are also more susceptible to spontaneous tumor development (Zanesi et al. 2001).
Gene therapy studies established that Fhit re-expression not only prevents development of cancer, but also contributes to tumor regression via apoptosis. Delivery of Fhit via adenovirus (Ad-Fhit) or adeno-associated virus (AAV-Fhit) significantly inhibits tumor development after NMBA treatment (Dumon et al. 2001). Immunohistochemical staining of tissue resected from sacrificed animals showed opposite Bax and Bcl2 staining in control versus Fhit-expressing animals, suggesting that failure to induce apoptosis contributed to tumor development in control animals while apoptosis due to delivery of Fhit in vector-treated animals contributed to tumor regression (Ishii et al. 2003).
Fhit Encodes a Dinucleoside Polyphosphate Hydrolase
Fhit and its homologs constitute one branch of histidine triad (HIT) proteins that have been strongly conserved throughout evolution. HIT proteins encode enzymes that bind and hydrolyze or phosphorylyze unusual nucleotide substrates. The HIT motif (His-Ø-His-Ø-His-Ø-Ø, in which Ø is a hydrophobic amino acid) forms part of the active site (Brenner 2002). Fhit protein and its orthologs are dinucleoside polyphosphate hydrolases, enzymes which bind and cleave a class of dinucleotides, typically containing two adenosines joined by 5'-5' bridges of 3–5 phosphates. The favored substrate of Fhit is ApppA (Barnes et al. 1996) and the substrate typically used in vitro is GpppBODIPY (Draganescu et al. 2000). Hydrolysis of such compounds produces a nucleoside monophosphate product (e.g., AMP or GMP) plus the other nucleotide or analog (e.g., ADP or ppBODIPY).
The Fhit Protein Interactome
Protein association studies have attempted to place Fhit into specific cellular pathways that might depend on diadenosine polyphosphates for signaling. Both wild-type Fhit and the catalytically inactive H96N mutant were shown to interact with tubulin in vitro and promote microtubule assembly in the presence of the microtubule binding proteins MAP 2 and Tau (Chaudhuri et al. 1999). In addition, yeast two-hybrid data suggest that Fhit interacts with the SUMO conjugating enzyme, Ubc9 (Shi et al. 2000). Biochemical assays have confirmed an interaction in vitro with one report (Golebiowski et al. 2004) but not another (Shi et al. 2000) indicating ApppA-dependent protein binding. Another appealing model is that Fhit may bind and destablize Mdm2, an E3 ubiquitin ligase that marks p53 for degradation (Nishizaki et al. 2004).
Fhit and a member of the nitrilase superfamily are encoded as fusion proteins in worms and flies (Pekarsky et al. 1998), suggesting that Fhit and Nit1, the ortholog of the nitrilase-related domain may participate in the same pathway and associate as seen in the crystal structure of the worm NitFhit protein (Pace et al. 2000). The data suggest that like Fhit, Nit1 functions as a tumor suppressor. The Nit1 knockout mouse shows increased susceptibility to NMBA-induced carcinogenesis and increased cell growth in culture (Semba et al. 2006). Moreover, loss of Fhit and loss of Nit1 are apparently additive for tumor suppression (Sun et al. 2009).
The Holy Grail in the molecular oncology of Fhit would be a set of protein interactions that provide a mechanism by which epithelial cells acquiring inactivating mutations in FHIT would obtain a survival advantage. The mechanism would also have to account for how Fhit re-expression induces apoptosis in a manner that depends on diadenosine polyphosphate binding. Though it is almost axiomatic that no unifying theory is accepted, there is one proposed mechanism notable for its far-reaching scope. According to this mechanism, Fhit transits to mitochondria via interaction with Hsp60 where it interacts with and stabilizes ferredoxin reductase (Fdxr) (Trapasso et al. 2008). Fdxr transfers electrons from NADPH to cytochrome P450 via ferredoxin. Under stress, Fdxr levels increase, leading to depletion of NADPH, which is required to detoxify reactive oxygen species (ROS), thereby inducing ROS-mediated apoptosis. Consistent with the expectation that diadenosine polyphosphate binding is limiting for the tumor suppressing function of Fhit, the investigators have shown that mutations that reduce substrate binding also reduce mitochondrial localization, association with Hsp60 and Fdxr, and cellular responses to oxidative damage (Pichiorri et al. 2009).
Genetic analysis of Fhit has indicated that reintroduction of this protein induces apoptosis in cancer cells with FHIT deletions, which are a common cancer genotype. Challenges remain in defining the specific consequences of Fhit inactivation in the preneoplastic epithelial cells from which Fhit is typically lost and in defining the relationship between diadenosine polyphosphate signaling and cellular survival in cells, early in the process of carcinogenesis.
- Nishizaki M, Sasaki J, Fang B, Atkinson EN, Minna JD, Roth JA, et al. Synergistic tumor suppression by coexpression of FHIT and p53 coincides with FHIT-mediated MDM2 inactivation and p53 stabilization in human non-small cell lung cancer cells. Cancer Res. 2004;64(16):5745–52.PubMedCrossRefGoogle Scholar