- Email: [email protected]
Roles of FHIT and WWOX fragile genes in cancer
Cancer Letters 232 (2006) 27–36 www.elsevier.com/locate/canlet
Mini Review
Roles of FHIT and WWOX fragile genes in cancer Dimitrios Iliopoulosa, Gulnur Gulerb, Shuang-Yin Hanc, Teresa Drucka, Michelle Otteyd, Kelly A. McCorkella,d, Kay Huebnera,* a
Department of Molecular Virology, Immunology and Medical Genetics, Comprehensive Cancer Center, Ohio State University, Columbus, OH, USA b Department of Pathology, Hacettepe University, Ankara, Turkey c Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA, USA d School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA. Received 30 May 2005; accepted 6 June 2005
Abstract It was hypothesized as early as 1986, that the recently discovered common fragile sites could facilitate recombination events, such as deletions and translocations, that result in clonally expanded cancer cell populations with specific chromosome alterations in specific cancer types. A natural extension of this hypothesis is that the clonal expansion must be driven by alteration of genes at recombination breakpoints whose altered functions actually drive clonal expansion. Nevertheless, when the FHIT gene was discovered at FRA3B, the most active common chromosome fragile region, and proposed as an example of a tumor suppressor gene altered by chromosome translocations and deletions, a wave of reports suggested that the FHIT gene was altered in cancer simply because it was in a fragile region and not because it had contributed to the clonal expansion, thus turning the original hypothesis upside down. Now, after nearly ten years and more than 500 FHIT reports, it is apparent that FHIT is an important tumor suppressor gene and that there are genes at other fragile regions that contribute significantly to development of cancer. A second fragile gene with a demonstrated role in cancer development is the WWOX gene on chromosome 16q; alterations to the WWOX gene contribute to development of hormone responsive and other cancers. Results of our recent studies of these two fragile tumor suppressor genes were summarized at the first Fragilome meeting in Heidelberg, Feb. 2005. q 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: FHIT; WWOX; Fragile genes; Tumor suppressors; Breast cancer
1. Introduction
* Corresponding author. Address: 410 W 12th Ave, Wiseman Hall, Rm 455c, Ohio State University Comprehensive Cancer Center, Columbus, OH 43210, USA. Tel.: C1 614 292 4850; fax: C1 614 292 3312. E-mail address: [email protected] (K. Huebner).
Eighty common and 24 rare fragile sites, encompassing more than 100 Mb of DNA are currently listed in the human genome database. Common or constitutive fragile sites are chromosome regions, observed in metaphase chromosomes of all
0304-3835/$ - see front matter q 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.06.048
28
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
individuals tested; that are more susceptible to breakage, rearrangements and deletions than other sites of the genome. These sites appear in metaphase chromosomes as gaps or breaks triggered by treatment of cells with agents that delay replication, such as the DNA polymerase inhibitor aphidicolin (for review [1]). What these gaps represent at the molecular level is not known, but it was recently suggested that they represent single-stranded regions that have escaped notice by the ATR-CHK1 DNA damage checkpoint and have thus persisted into mitosis [2]. Shortly after the discovery of common fragile sites, it was observed that the cytogenetic locations of these fragile sites frequently coincided with the cytogenetic locations nonrandomly altered in specific cancers [3]. Thus, it was possible that the recombinogenicity of fragile regions predisposed them to chromosome rearrangements during cancer development. If a specific chromosome rearrangement is observed consistently in clonally expanded preneoplasias or neoplasias, it is likely that the genetic change resulting from the rearrangement has contributed to the clonal expansion. Thus, a number of investigators supposed that there must be genes at some common fragile regions, that when deleted, translocated or amplified due to fragile site instability, contribute to initiation or progression of cancer. Several laboratories thus cloned and characterized portions of the genomic regions in the vicinity of the most active fragile site, FRA3B [4–6] and sought to understand the relationship between fragility and specific DNA sequences. Even before common fragile sites were observed, a familial, balanced chromosome translocation involving chromosome regions 3p14.2 and 8q24, was reported in a family with early onset, bilateral and multifocal clear cell renal cancers [7]. Several fragile site laboratory groups were seeking to clone the t(3;8) translocation breakpoint, with the idea that a gene disrupted by the translocation break might be a tumor suppressor whose alteration had contributed to development of the familial kidney cancers [4–6]. The search culminated in 1996, in the discovery of the very large (1.6 Mb) FHIT gene that encompassed the fragile region, as well as the translocation breakpoint in intron 3 of the FHIT gene [8,9] (Fig. 1A). Cancer cell lines and primary cancers were shown to harbor deletions within one or both FHIT alleles [8–10], Ref. [1] for review) and large fractions of various
types of cancers, including upper gastrointestinal tract, colon, cervical, lung and breast cancers, expressed reduced or no Fhit protein (for review [11]). Although in the early years after discovery of the FHIT gene there was considerable resistance to acceptance of FHIT among the galaxy of Tumor Suppressor Genes [12,13] by early 2005 there were more than 500 reports concerning alteration of the FHIT gene or protein in cancers arising in almost every organ of the body. Nine years after its discovery, through contributions of many researchers from all over the globe, the role of FHIT as the archetypal fragile tumor suppressor gene is well established and these studies are serving as models for the isolation and characterization of genes at other common human fragile sites.
Fig. 1. Schematic of the FHIT and WWOX genes. The chromosome arm where each gene is located is shown on the left. The chromosome band where the genes are located is magnified and the genomic structure of each gene is shown on the right. The dark boxes represent protein coding exons and the open boxes non amino acid coding exons. Many of the deletions detected in tumor cell lines are centered on FHIT exon 5 and WWOX exon 8.
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
2. FHIT is down-regulated in most cancers Cancer cell lines and primary cancers exhibiting hemi or homozygous deletions with endpoints within the FHIT gene and reduced or absent Fhit expression have been reported (for review [1]). Reduced expression of Fhit in pre-invasive lesions is also observed. In esophageal cancer, Mori et al., reported that most of the in-situ lesions, 50% of severe and moderate dysplasias and 33% of mild dysplasias were Fhit negative [14]; in the study of Kitamura et al. [15] reduced Fhit expression was seen in 68% of in situ and 43.5% of esophageal dysplastic lesions. Hao et al. [16] found reduced Fhit expression only in a small fraction of adenomatous colon lesions, but reduced Fhit expression was associated with greater degree of dysplasia. In cervical cancer Connolly et al. [17] observed reduced or absent Fhit staining in 71% of invasive cancers and in 52% of high-grade intraepithelial lesions (HSILs) with invasive cancer. In w85% of bronchial dysphasia’s there was loss of Fhit expression [18]. In our study of DCIS (ductal carcinoma in situ), reduced Fhit expression was observed in: 70% of pure DCIS and 52% of DCIS adjacent-to-invasive tumor cases. 20% of pure DCIS cases exhibited individual glands of adjacent normal tissue with absence of expression [19]. LOH at 3p14.2 alleles has been found in pre-invasive bronchial lesions and the non-neoplastic airway cells of smokers and former smokers [11] for review). These clinical findings supported the proposal that FHIT inactivation occurs in early steps of carcinogenesis in many organs. The association between exposure to carcinogens and FHIT inactivation was first noted in lung cancers, which are among the cancers with the highest frequency of complete loss of Fhit expression [18], suggesting that alteration of the FHIT gene through damage to the associated fragile region by environmental carcinogens, contributes substantially to the human cancer burden. In support of this idea, FHIT inactivation was observed twice as often in lung tumors of smokers (75%) relative to non-smokers (39%) [18,20]. Exposure to asbestos and to gamma irradiation during the Chernobyl accident caused an increase in FHIT inactivation in lung cancer and preneoplastic bronchial lesions [20,21]. Smoking
29
history and alcohol abuse associated with higher frequency of loss of Fhit expression was also reported in esophageal cancer [14]. If FHIT is one of the first targets of carcinogens, the ability of the host to repair this initial damage or to eliminate cells carrying damage to the FHIT locus, may prevent clonal expansion. In support of this idea, loss of FHIT function is observed more frequently in cancers developing in individuals with constitutional alterations to genes involved in DNA repair, such as the BRCA1 gene and mismatch repair genes [22–24].
3. Epigenetic mechanisms contribute to loss of Fhit expression in cancers Epigenetic changes to chromatin, such as DNA methylation and modifications in histone tails regulate transcription of numerous tumor suppressor genes (for review [25]). DNA hypermethylation of promoter regions of tumor suppressor genes has been reported to be associated with decreased gene expression. FHIT promoter hypermethylation associated with loss of gene expression was reported for lung, breast and esophageal cancer [26–28]. It has also been reported that in non-small cell lung and bladder carcinomas FHIT promoter hypermethylation correlated with worse prognosis; in lung carcinomas, it was reported to be a sign of worse prognosis, even in patients with earlier stage of disease [28,29]. Hypermethylation of FHIT was associated with pack-years of cigarette smoking in human lung squamous cell carcinomas [30,31]. In a study of rodent tumors, most showed reduced Fhit expression and methylation in promoter, exon 1 and intron 1 regions [32]. In rat primary tumors induced by different carcinogens, different CpG sites were methylated, suggesting tissue or carcinogenspecific methylation patterns; in a follow-up study, it was determined that Fhit methylation patterns in benign and malignant tumors induced by two different carcinogens (MNU, DMBA) were tissue, not carcinogen, specific [33]. In human cancers we observed differential patterns of FHIT methylation in neoplastic vs. non neoplastic tissues; FHIT intron 1 methylation may serve as a marker of early steps in breast cancer development, suggesting that targeted methylationspecific PCR (MSP) amplification would be useful in following treatment or prevention protocols [28].
30
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
4. Fhit deficient mouse models to study cancer development, prevention and therapy Because the murine Fhit gene encompasses a common fragile site, Fra 14A2 and homozygous deletions of Fhit exons were observed in two mouse cancer lines (reviewed in [11]), the mouse model was chosen to study Fhit inactivation effects. FhitK/K mice, although being healthy and fertile, showed increased susceptibility to spontaneous and carcinogen-induced tumors [34]. All Fhit C/K mice developed several forestomach and sebaceous gland tumors after six doses of intragastric N-nitrosomethylbenzylamine (NMBA) while only 25% of wildtype mice developed tumors. The tumors were a mixture of benign, in-situ and invasive lesions. The NMBAinduced tumor spectrum in FhitC/K mice was similar to a human syndrome called Muir-Torre syndrome, a variant of hereditary non-polyposis colorectal cancer, which is often caused by inactivation of a mismatch repair gene, usually MSH2. Homozygous deletions of FHIT were observed in half of the cancer cell lines with a mutant mismatch repair gene; among nine Msh2-negative human colon cancer cells eight were negative for Fhit, and Fhit loss was reported in 90% of BRCA1 and 2 associated cancers [22–24]. These correlations suggest that fragile genes are especially vulnerable to damage-induced alterations in cells with ‘caretaker’ gene deficiencies. After showing increased frequency of tumors in Fhit deficient mice, the next question was: can FHIT gene therapy prevent and reverse development of these tumors (for review, [35]). Esophageal cancer cell lines transfected with Adeno-FHIT virus exhibited reduced proliferation and less tumorigenic behaviour when injected into immunodeficient mice. The growth of xenografts of poorly differentiated pancreatic cancer cell lines in nude mice was suppressed by Adeno-FHIT and AAV-FHIT (adenoassociated-virus-containing FHIT infection). AdenoFHIT infection was also associated with increased survival of mice with disseminated pancreatic cancer. AAV-FHIT or Adeno-FHIT treated Fhit knockout mice developed significantly fewer and smaller tumors following NMBA exposure, suggesting that local delivery of the FHIT gene to human premalignant lesions might be an effective preventive strategy. Since FHIT is one of the first genes damaged in
response to carcinogen exposure, for example in normal and metaplastic respiratory epithelium of smokers, and is frequently inactivated in tumors of repair deficient individuals, FHIT gene therapy offers promise for treatment and prevention of various types of cancers.
5. Fhit protein function Despite strong evidence of Fhit tumor suppressor function, our knowledge of specific Fhit signal pathways and mechanisms involved in the suppressor activity is limited. It is well-known that overexpression of Fhit results in apoptosis in Fhit deficient cancer cells (for review [35]), and recent studies have demonstrated a role for Fhit in the responses to genotoxic damage induced by UVC light, mitomycin C, camptothecin and ionizing radiation [36,37]; w10-fold more colonies of Fhit-deficient cells survived exposure to high UVC doses (Fig. 2). After mitomycin C treatment w6-fold and after UVC treatment 3.5-fold more Fhit-positive human cancer cells than Fhit-negative cells had died and UVC surviving FhitK/K cells showed more than 5-fold increased mutation frequency. The histidine triad motif –HxHxH– of Fhit is conserved in all species. Fhit protein is homologous to S. pombe Aph1 enzyme, which has diadenosine triphosphate hydrolase activity, an enzymatic function that is conserved from yeast to human [1]. However, it has been shown that ‘hydrolase dead’ Fhit with the central histidine mutated to asparagine (H96N) suppresses tumorigenity as well as wild type protein. Structural studies show that binding of the Fhit dimer with two molecules of diadenosine triphoshate, results in highly phosphorylated surfaces, which may have signaling activity [38]. Very recently it was discovered that Fhit can be phosphorylated by Src and Src family kinases in vivo and in vitro at amino acid Y114 [39]. Because Fhit is a dimer, there are three forms of Fhit, unphosphorylated, monophosphorylated, with only one monomer of the dimer phosphorylated, and the diphosphorylated form. The phosphorylated Fhit monomer is upshifted on PAGE gels, so that phosphorylation is easily followed on immunoblots. Phosphorylation of endogenous Fhit is not observed in cells in tissue culture, thus the consequences of phosphorylation of the Y114
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Fig. 2. FhitK/K and C/C mouse kidney cells were seeded, exposed to 60 J/m2 UV, collected, counted and re-seeded. After 18 days the plates were fixed and stained with Giemsa, and colonies counted. FhitK/K cells formed w90 colonies and FhitC/C cells formed w10 colonies. B. MKN-74 E4 and MKN-74 A66 human gastric cancer cells were seeded, exposed to 60 J/m2 UV, collected, counted, re-seeded and allowed to form colonies. After 18 days plates were fixed and stained with Giemsa. The colony numbers for the E4 (Fhit negative cells) were about ten times greater than for A66 (Fhit positive) cells: E4 w300 colonies. A66 w30 colonies. This figure was first published as Fig. 3b in Ref. [36] (British J Cancer).
tyrosine residue has been difficult to examine in vitro. Mutant Fhit constructs, with Y114 and nearby amino acids altered, have been prepared in plasmids as well as Adenovirus vectors, in order to study the biological function of Fhit, with and without phosphorylation at Y114.
6. The WWOX tumor suppressor gene at the FRA16D fragile site 6.1. WWOX structure and expression in cancer The WWOX gene spans a genomic region of more than one million nucleotide base pairs located at 16q23.3-24.1, a chromosome region commonly
31
involved in LOH in many different types of cancer (Fig. 1B); it is composed of nine exons, encoding a cDNA of 1245 bp [40]. Studies of esophageal squamous cell carcinoma, non-small cell lung cancer and breast cancer showed a high LOH rate, low mutation rate and expression of aberrant transcripts of the WWOX gene (reviewed in [41]). Northern blot and RT-PCR analyses from breast and ovarian cancer cells revealed the presence of transcripts of smaller size, representing abnormally spliced versions of WWOX [40,43]. WWOX mRNA expression is reduced in a series of human breast tumors and breast cancer cell lines [41–45]. In the first study of Wwox expression in clinical cancer tissues, coordinate absence or reduction of Fhit and Wwox expression in 55 and 63% of invasive breast tumors was reported (examples in Fig. 3) [46]. Reduced Wwox staining was found more frequently in ER (K) or scanty positive tumors. In adjacent normal tissue, reduced Wwox staining was observed in 32.9% of cases. Of all cases that showed Wwox staining in less than 10% of normal breast tissue were postmenopausal or exposed to neoadjuvant chemotherapy, implying that Wwox expression may be associated with the level of steroid hormone expression and can be effected by chemotherapeutic drugs. Nunez et al. [47] also reported frequent loss of Wwox expression in association with decreased ER expression in breast cancer. In our study of DCIS, reduced Fhit and Wwox expression were also observed in: (a) 70 and 68% of pure DCIS; (b) 52 and 55% of DCIS adjacent-to-invasive tumor cases; (c) some individual glands of adjacent normal tissue of 20 and 50% of pure DCIS cases [19]. Reduced Wwox expression in adjacent normal tissue was observed in 30% of DCIS adjacent-to-invasive cases. Reduced Fhit and Wwox expression was observed in 61% of adjoining invasive tumors. In all normal, pure DCIS and DCIS adjacent-to-invasive lesions, Fhit and Wwox expression were positively associated. Fhit and Wwox were more frequently reduced in high-grade lesions in the DCIS adjacentto-invasive group. In summary, Fhit and Wwox were lost coordinately in in-situ breast cancer, losses that may contribute to the high grade DCIS-invasive tumor pathway. Frequent loss and downregulation of Wwox expression was reported in human hepatocellular carcinoma cell lines; in primary gastric cancers,
32
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Fig. 3. Immunohistochemical staining of Fhit and Wwox. (A) An example of strong Fhit staining intensity in tumor and normal luminal epithelial cells (between arrows). (B) Loss of Fhit staining in tumor cells. (C) Similarly reduced Wwox staining intensity in adjacent normal luminal epithelial cells and tumor tissue. (D) Reduced Wwox staining intensity of tumor cells (adjacent normal tissue is between arrows). (E) Reduced Wwox staining intensity in normal luminal epithelial cells. (F) Intense cytoplasmic staining of Wwox in !10% of neoplastic cells. Original magnification !100 (B,F); !200 (C–E); !400 (A). (The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinomas, Gulnur Guler, CANCER, 100(8): 1605–14, Wiley–Liss, Inc.).
LOH in WWOX locus was observed in 31% and loss of Wwox protein expression in 65% of cases and in the same study a high correlation between Fhit and Wwox expression levels was found (reviewed in [41]).
7. Epigenetic mechanisms regulate WWOX expression Expression of many tumor suppressor genes is down-regulated in cancer by epigenetic mechanisms;
Fig. 4. Wwox regulatory region. The highest percentage of CpG sites starts from K183 to C177 relative to the transcription initiation site. 3 CpGs in the promoter region correspond to Sp1 binding sites.
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
CpG rich areas (islands) of gene promoter regions are frequently methylated in cancer (for review, [25]), and histone modifications, such as deacetylation and lysine methylation, change the chromatin structure in the promoter region of tumor suppressor genes, so that transcription factors cannot bind and transcription is blocked. Significantly, modifications are reversible by demethylating agents and HDAC inhibitors and are potential targets for cancer therapy. We inspected an interval including 1 kb upstream and downstream from the WWOX transcription initiation site for in silico CpG island identification [48]. A region 406 bp upstream of the WWOX transcription initiation site, exon 1 and part of intron 1 (the first 125 bp) exhibited 66% CG residues, the region from K183 to C177 has the highest percentage (w70%) of CpG sites (Fig. 4). Ishii et al., reported a possible role of epigenetic mechanisms in WWOX transcriptional regulation [49]. Specifically treatment of K562 leukemia cells with 5-Aza-2-deoxycytidine (demethylating agent) and depsipeptide (HDAC inhibitor) increased WWOX mRNA expression. Kuroki et al. [50] studied WWOX promoter methylation status in pancreatic cancer and reported methylation in the region K148 to K37 respective to the transcription initiation site (C1) in pancreatic adenocarcinomas; WWOX promoter methylation status was associated with its expression and after treatment with 5-Aza-deoxycytidine in Hs766T cells, Wwox was re-expressed. We have been interested in mechanisms that control WWOX fragile gene expression and inactivation in human cancers and have assessed lung, breast and bladder cancers for control of expression by methylation of 5 0 CpG islands. Frozen lung and breast tumor tissue, adjacent non-neoplastic and normal tissue samples were used in this study [48]. Wwox expression was reduced in breast and lung cancers in association with hypermethylation. Differential patterns of WWOX methylation were observed in neoplastic vs adjacent non-neoplastic tissues, suggesting that targeted MSP amplification could be useful in following treatment or prevention protocols. WWOX promoter MSP differentiates DNA of lung cancer from DNA of adjacent lung tissue. WWOX and FHIT promoter methylation is detected in tissue adjacent to breast cancer and WWOX exon 1 MSP distinguishes breast cancer DNA from DNA of
33
adjacent and normal tissue. Differential methylation in cancerous vs adjacent tissues suggests that WWOX hypermethylation analyses could enrich a panel of DNA methylation markers.
8. Wwox protein structure and function Bednarek et al., described the WWOX gene at 16q23.3-24.1 and observed that Wwox protein contains two WW domains in the NH2 terminus and a SDR (short chain dehydrogenase/reductase) domain [40]. WW domains are globular domains consisting of w40 amino acids, of which two tryptophans and an invariant proline are highly conserved [51]. Like SH3 domains, the WW domains are characterized by interactions with proline-containing ligands and mediate protein–protein interactions. WW domains can be grouped into four classes according to their ligand binding preferences and recently it was suggested that WW domain interactions can be modulated by tyrosine phosphorylation [52]. The short-chain dehydrogenase reductase (SDR) family includes steroid dehydrogenase enzymes and more than 60 other proteins from human, mammalian, insect, and bacterial sources. Most family members contain tyrosine and lysine of the catalytic triad in a YxxxK sequence. X-ray crystal structures of 13 members of the family showed that when the alphacarbon backbone of the cofactor binding domains of the structures are superimposed, the conserved residues are at the core of the structure and in the cofactor binding domain, but not in the substrate binding pocket [53]. The SDR domain in Wwox protein may play an important role in sex hormone regulated cancers such as breast, ovarian and prostate cancer. Chang et al. [54] recently reported that Wwox is upregulated in COS7 fibroblasts and DU145 prostate cancer cells after 17-b-estradiol treatment. WWOX spans the common fragile site FRA16D and it has been reported that FRA16D and the corresponding gene, WWOX, are highly conserved in the mouse at Fra8E1 (for review [41]). In studies seeking the function of Wwox, Ludes-Meyers et al. [55] showed that Wwox contains a Group I WW domain that binds known cellular proteins containing the specific ligand PPXY.
34
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Aqeilan et al., [56] demonstrated a physical interaction between the first WW domain of Wwox and p73, a p53 homolog. The PPPY motif in the WW1 domain can be phosphorylated by Src-family tyrosine kinases, perhaps indicating that Src directly phosphorylates this residue. Overexpression of Wwox caused redistribution of p73 from the nucleus to the cytoplasm and suppression of p73 mediated interactions. Wwox also binds to the PPPY motif of AP2g via its first WW domain [57]. Ap2g encodes a transcription factor and is frequently up-regulated in breast carcinomas. Wwox overexpression triggered redistribution of nuclear Ap2g to the cytoplasm, hence suppressing its transactivating function. Interestingly, point mutation in the terminal tyrosine of the PPPY motif abrogates Wwox-p73 and -Ap2g interactions, emphasizing the principal of sequence-specific protein–protein interactions as a determinant of precise biological outputs. Thus Wwox appears to act as a modulator of transcription by binding these proteins. Our limited information about Wwox function suggests that, like Fhit, Wwox may participate in apoptotic pathways, though details of the Wwox signal pathways are not yet known.
9. Conclusions The apparent coordinate inactivation of the two fragile genes, FHIT and WWOX, in breast cancer, at the protein expression and promoter methylation level, [19,33], suggests the possibility of coordinate carcinogen damage to multiple fragile sites in some types of cancer. As we learn more about the fragile sites and the functions of genes encompassing them, we will gain further insight into their roles in initiation and progression of carcinogenesis.
Acknowledgements This work was supported by DOD predoctoral grant BC043090 (D.I), support from Hacettape University and the National Cancer Institute, USA (G.G), T32-HLO7780 (K.A.M) and the NCI grants CA77738 and CA56036.
References [1] K. Huebner, C.M. Croce, FRA3B and other common fragile sites: the weakest links, Nat. Rev. Cancer 1 (2001) 214–221. [2] A.M. Casper, P. Nghiem, M. Arlt, T.W. Glover, ATR regulates fragile site stability, Cell 111 (2002) 779–789. [3] J. Yunis, A.L. Soreng, Constitutive fragile sites and cancer, Science 226 (1984) 1199–1204. [4] F.L. Boldog, B. Waggoner, T.W. Glover, I. Chumakov, D. Le Paslier, D. Cohen, et al., Integrated YAC contig containing the 3p14.2 hereditary renal carcinoma 3;8 translocation breakpoint and the fragile site FRA3B, Genes Chromosomes Cancer 11 (1994) 216–221. [5] C.M. Wilke, S.W. Guo, B.K. Hall, F. Boldog, R.M. Gemmill, S. Chandrasekharappa, et al., Multicolor FISH mapping of YAC clones in 3p14 and identification of a YAC spanning both FRA3B and the t(3;8) associated with hereditary renal cell carcinoma, Genomics 22 (1994) 319–326. [6] W. Paradee, C.M. Wilke, L. Wang, R. Shridhar, C.M. Mullins, A. Hoge, et al., A 350-kb cosmid contig in 3p14.2 that crosses the t(3;8) hereditary renal cell carcinoma translocation breakpoint and 17 aphidicolin-induced FRA3B breakpoints, Genomics 35 (1996) 87–93. [7] A.J. Cohen, F.P. Li, S. Berg, D.J. Marchetto, S. Tsai, S.C. Jacobs, R.S. Brown, Hereditary renal-cell carcinoma associated with a chromosomal translocation, N. Engl. J. Med. 301 (1979) 592–595. [8] M. Ohta, H. Inoue, M.G. Cotticelli, K. Kastury, R. Baffa, J. Palazzo, et al., The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers, Cell 84 (1996) 587–597. [9] D. Zimonjic, T. Druck, M. Ohta, K. Kastury, C.M. Croce, N.C. Popescu, K. Huebner, Positions of chromosome 3p14.2 fragile sites (FRA3B) within the FHIT gene, Cancer Res. 57 (1997) 1166–1170. [10] G. Sozzi, S. Tornielli, E. Tagliabue, L. Sard, F. Pezzella, U. Pastorino, et al., Absence of Fhit protein in primary lung tumors and cell lines with FHIT gene abnormalities, Cancer Res. 57 (1997) 5207–5212. [11] K. Huebner, C.M. Croce, Cancer and the FRA3B/FHIT fragile locus: it’s a HIT, Br. J. Cancer 10 (2003) 1501–1506. [12] S. Thiagalingam, N.A. Lisitsyn, M. Hamaguchi, M.H. Wigler, J.K. Willson, S.D. Markowitz, et al., Evaluation of the FHIT gene in colorectal cancers, Cancer Res. 56 (1996) 2936–2939. [13] M.M. Le Beau, H. Drabkin, T.W. Glover, R. Gemmill, F.V. Rassool, T.W. McKeithan, D.I. Smith, An FHIT tumor suppressor gene? Genes Chromosomes Cancer 21 (1998) 281–289. [14] M. Mori, K. Mimori, T. Shiraishi, H. Alder, H. Inoue, Y. Tanaka, et al., Altered expression of Fhit in carcinoma and precarcinomatous lesions of the esophagus, Cancer Res. 60 (2000) 1177–1182. [15] A. Kitamura, K. Yashima, E. Okamoto, H. Andachi, A. Hosoda, Y. Kishimoto, et al., Reduced Fhit expression
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25] [26]
[27]
[28]
[29]
occurs in the early stage of esophageal tumorigenesis: no correlation with p53 expression and apoptosis, Oncology 61 (2001) 205–211. X.P. Hao, J.E. Willis, T.G. Pretlow, J.S. Rao, G.T. MacLennan, I.C. Talbot, T.P. Pretlow, Loss of fragile histidine triad expression in colorectal carcinomas and premalignant lesions, Cancer Res. 60 (2000) 18–21. D.C. Connolly, D.L. Greenspan, R. Wu, X. Ren, R.L. Dunn, K.V. Shah, et al., Loss of Fhit expression in invasive cervical carcinomas and intraepithelial lesions associated with invasive disease, Clin. Cancer Res. 6 (2000) 3505–3510. G. Sozzi, U. Pastorino, L. Moiraghi, E. Tagliabue, F. Pezzella, C. Ghirelli, et al., Loss of FHIT function in lung cancer and preinvasive bronchial lesions, Cancer Res. 58 (1998) 5032– 5037. G. Guler, A. Uner, N. Gu¨ler, S.Y. Han, D. Iliopoulos, P. McCue, K. Huebner, Concordant loss of fragile gene expression early in breast cancer development, Pathol. Int. 8 (2005) 471–478. L. Pylkkanen, H. Wolff, T. Stjernvall, P. Tuominen, T. Sioris, A. Karjalainen, et al., Reduced Fhit protein expression and loss of heterozygosity at FHIT gene in tumours from smoking and asbestos-exposed lung cancer patients, Int. J. Oncol. 20 (2002) 285–290. V. Chizhikov, S. Chikina, A. Gasparian, I. Zborovskaya, E. Steshina, G. Ungiadze, et al., Molecular follow-up of preneoplastic lesions in bronchial epithelium of former Chernobyl clean-up workers, Oncogene 21 (2002) 2398–2405. S. Ingvarsson, B.A. Agnarsson, B.I. Sigbjornsdottir, J. Kononen, O.P. Kallioniemi, R.B. Barkardottir, et al., Reduced Fhit expression in sporadic and BRCA2-linked breast carcinomas, Cancer Res. 59 (1999) 2682–2689. M. Mori, K. Mimori, T. Masuda, K. Yoshinaga, K. Yamashita, A. Matsuyama, H. Inoue, Absence of Msh2 protein expression is associated with alteration in the FHIT locus and Fhit protein expression in colorectal carcinoma, Cancer Res. 61 (2001) 7379–7382. B.C. Turner, M. Ottey, D.B. Zimonjic, M. Potoczek, W.W. Hauck, E. Pequignot, et al., The fragile histidine triad/common chromosome fragile site 3B locus and repairdeficient cancers, Cancer Res. 62 (2002) 4054–4060. C.L. Hsieh, P.A. Jones, Meddling with methylation, Nat. Cell Biol. 5 (2003) 502–504. H. Tanaka, Y. Shimada, H. Harada, M. Shinoda, S. Hatooka, M. Imamura, K. Ishizaki, Methylation of the 5 0 CpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas, Cancer Res. 58 (1998) 3429–3434. S. Zochbauer-Muller, K.M. Fong, A. Maitra, S. lam, J. Goradts, R. Ashfuq, et al., 5 0 CpG island methylation of the FHIT gene is correlated with loss of gene expression in lung and breast cancer, Cancer Res. 61 (2001) 3581–3585. R. Maruyama, K. Sugio, I. Yoshino, Y. Maehara, A.F. Gazdar, Hypermethylation of FHIT as a prognostic marker in nonsmall cell lung carcinoma, Cancer 100 (2004) 1472–1477. R. Maruyama, S. Toyooka, K.O. Toyooka, K. Harada, A.K. Virmani, S. Zochbauer-Muller, et al., Aberrant promoter
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
35
methylation profile of bladder cancer and its relationship to clinicopathological features, Cancer Res. 61 (2001) 8659– 8663. J.S. Kim, H. Kim, Y.M. Shim, J. Han, J. Park, D.H. Kim, Aberrant methylation of the FHIT gene in chronic smokers with early stage squamous cell carcinoma of the lung, Carcinogenesis 25 (2004) 2165–2171. H. Kim, Y.M. Kwon, J.S. Kim, H. Lee, J.H. Park, Y.M. Shim, et al., Tumor-specific methylation in bronchial lavage for the early detection of non-small-cell lung cancer, J. Clin. Oncol. 22 (2004) 2363–2370. S.Y. Han, D. Iliopoulos, T. Druck, G. Guler, C.J. Grubbs, M. Pereira, et al., CpG methylation in the Fhit regulatory region: relation to Fhit expression in murine tumors, Oncogene 23 (2004) 3990–3998. G. Guler, D. Iliopoulos, S.Y. Han, L.Y. Fong, R.A. Lubet, C.J. Grubbs, K. Huebner, Hypermethylation patterns in the Fhit regulatory region are tissue specific, Mol. Carcinog. 3 (2005) 175–181. N. Zanesi, V. Fidanza, L.Y. Fong, R. Mancini, T. Druck, M. Valtieri, et al., The tumor spectrum in FHIT-deficient mice, Proc. Natl. Acad. Sci. USA 98 (2001) 10250–10255. H. Ishii, K.R. Dumon, A. Vecchione, L.Y. Fong, R. Baffa, K. Huebner, C.M. Croce, Potential cancer therapy with the fragile histidine triad gene: review of the preclinical studies, J. Am. Med. Assoc. 19 (2001) 2441–2449. M. Ottey, S.Y. Han, T. Druck, B.L. Barnoski, K.A. McCorkell, C.M. Croce, et al., Fhit-deficient normal and cancer cells are mitomycin C and UVC resistant, Br. J. Cancer 9 (2004) 1669–1677. B. Hu, S.Y. Han, X. Wang, M. Ottey, M.B. Potoczek, A. Dicker, et al., Involvement of the Fhit gene in the ionizing radiation-activated ATR/CHK1 pathway, J. Cell Physiol. 202 (2005) 518–523. H.C. Pace, P.N. Garrison, A.K. Robinson, L.D. Barnes, A. Draganescu, A. Rosler, et al., Genetic, biochemical, and crystallographic characterization of Fhit-substrate complexes as the active signaling form of Fhit, Proc. Natl. Acad. Sci. USA 95 (1998) 5484–5489. Y. Pekarsky, P.N. Garrison, A. Palamarchuk, N. Zanesi, R.I. Aqeilan, K. Huebner, et al., Fhit is a physiological target of the protein kinase Src, Proc. Natl. Acad. Sci. USA 101 (2004) 3775–3779. A.K. Bednarek, K.J. Laflin, R.L. Daniel, Q. Liao, K.A. Hawkins, C.M. Aldaz, WWOX, a novel WW domaincontaining protein mapping to human chromosome 16q23.324.1, a region frequently affected in breast cancer, Cancer Res. 60 (2000) 2140–2145. J.H. Ludes-Meyers, A.K. Bednarek, N.C. Popescu, M. Bedford, C.M. Aldaz, WWOX, the common chromosomal fragile site, FRA16D, cancer gene, Cytogenet. Genome Res. 100 (2003) 101–110. A.K. Bednarek, C.L. Keck-Waggoner, R.L. Daniel, K. Laflin, P. Bergsagel, K. Kiguchi, et al., FRA16D gene, behaves as a suppressor of tumor growth, Cancer Res. 61 (2001) 8068–8073. A.J. Paige, K.J. Taylor, A. Stewart, J.G. Sgouros, H. Gabra, G.C. Sellar, et al., A 700-kb physical map of a region of
36
[44]
[45]
[46]
[47]
[48]
[49]
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36 16q23.2 homozygously deleted in multiple cancers and spanning the common fragile site FRA16D, Cancer Res. 60 (2000) 1690–1697. M. Mangelsdorf, K. Ried, E. Woollatt, S. Dayan, H. Eyre, M. Finnis, et al., Chromosomal fragile site FRA16D and DNA instability in cancer, Cancer Res. 60 (2000) 1683–1689. A.K. Bednarek, C.L. Keck-Waggoner, R.L. Daniel, K.J. Laflin, P.L. Bergsagel, K. Kiguchi, et al., FRA16D gene, behaves as a suppressor of tumor growth, Cancer Res. 61 (2001) 8068–8073. G. Guler, A. Uner, N. Guler, S.Y. Han, D. Iliopoulos, W.W. Hauck, et al., The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinoma, Cancer 8 (2004) 1605–1614. M.I. Nunez, J. Ludes-Meyers, M.C. Abba, H. Kil, N. Abbey, R.E. Page, et al., Frequent loss of WWOX expression in breast cancer: correlation with estrogen receptor status, Breast Cancer Res. Treat. 7 (2005) 32–1614. D. Iliopoulos, G. Guler, S.Y. Han, D. Johnston, T. Druck, K.A. McCorkell, et al., Fragile genes as biomarkers: epigenetic control of WWOX and FHIT in lung, breast and bladder cancer, Oncogene 9 (2005) 1625–1633. H. Ishii, K. Mimor, T. Inageta, Y. Murakumo, A. Vecchione, M. Mori, Y. Furukawa, Components of DNA damage checkpoint pathway regulate UV exposuredependent alterations of gene expression of FHIT and WWOX at chromosome fragile sites, Mol. Cancer Res. 3 (2005) 130–138.
[50] T. Kuroki, S. Yendamuri, F. Trapasso, A. Matsuyama, R.I. Aqeilan, H. Alder, et al., The tumor suppressor gene WWOX at FRA16D is involved in pancreatic carcinogenesis, Clin. Cancer Res. 10 (2004) 2459–2465. [51] M.J. Macias, WW and SH3 domains, two different scaffolds to recognize proline-rich ligands, FEBS Lett. 513 (2002) 30–37. [52] J.L. Ilsley, M. Sudol, S.J. Winder, The WW domain: linking cell signalling to the membrane cytoskeleton, Cell signaling 14 (2002) 183–189. [53] W.L. Duax, D. Ghosh, V. Pletnev, Steroid dehydrogenase structures, mechanism of action, and disease, Vitam Horm. 58 (2000) 121–148. [54] N.S. Chang, L. Schultz, L.J. Hsu, J. Lewis, M. Su, C.I. Sze, 17beta-Estradiol up-regulates and activates WOX1/WWOXv1 and WOX2/WWOXv2 in vitro: potential role in cancerous progression of breast and prostate to a premetastatic state in vivo, Oncogene 24 (2005) 714–723. [55] J.H. Ludes-Meyers, H. Kil, A.K. Bednarek, J. Drake, M. Bedford, C.M. Aldaz, W.W.O.X. binds, the specific proline-rich ligand PPXY: identification of candidate interacting proteins, Oncogene 23 (2004) 5049–5055. [56] R.I. Aqeilan, Y. Pekarsky, J.J. Herrero, A. Palamarchuk, J. Letofsky, T. Druck, Functional association between Wwox tumor suppressor protein and p73, a p53 homolog, Proc. Natl. Acad. Sci. USA 101 (2004) 4401–4406. [57] R.I. Aqeilan, A. Palamarchuk, R.J. Weigel, J.J. Herrero, Y. Pekarsky, C.M. Croce, Physical and functional interactions between the Wwox tumor suppressor protein and the AP2gamma transcription factor, Cancer Res. 64 (2004) 8256–8266.
Mini Review
Roles of FHIT and WWOX fragile genes in cancer Dimitrios Iliopoulosa, Gulnur Gulerb, Shuang-Yin Hanc, Teresa Drucka, Michelle Otteyd, Kelly A. McCorkella,d, Kay Huebnera,* a
Department of Molecular Virology, Immunology and Medical Genetics, Comprehensive Cancer Center, Ohio State University, Columbus, OH, USA b Department of Pathology, Hacettepe University, Ankara, Turkey c Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA, USA d School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA. Received 30 May 2005; accepted 6 June 2005
Abstract It was hypothesized as early as 1986, that the recently discovered common fragile sites could facilitate recombination events, such as deletions and translocations, that result in clonally expanded cancer cell populations with specific chromosome alterations in specific cancer types. A natural extension of this hypothesis is that the clonal expansion must be driven by alteration of genes at recombination breakpoints whose altered functions actually drive clonal expansion. Nevertheless, when the FHIT gene was discovered at FRA3B, the most active common chromosome fragile region, and proposed as an example of a tumor suppressor gene altered by chromosome translocations and deletions, a wave of reports suggested that the FHIT gene was altered in cancer simply because it was in a fragile region and not because it had contributed to the clonal expansion, thus turning the original hypothesis upside down. Now, after nearly ten years and more than 500 FHIT reports, it is apparent that FHIT is an important tumor suppressor gene and that there are genes at other fragile regions that contribute significantly to development of cancer. A second fragile gene with a demonstrated role in cancer development is the WWOX gene on chromosome 16q; alterations to the WWOX gene contribute to development of hormone responsive and other cancers. Results of our recent studies of these two fragile tumor suppressor genes were summarized at the first Fragilome meeting in Heidelberg, Feb. 2005. q 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: FHIT; WWOX; Fragile genes; Tumor suppressors; Breast cancer
1. Introduction
* Corresponding author. Address: 410 W 12th Ave, Wiseman Hall, Rm 455c, Ohio State University Comprehensive Cancer Center, Columbus, OH 43210, USA. Tel.: C1 614 292 4850; fax: C1 614 292 3312. E-mail address: [email protected] (K. Huebner).
Eighty common and 24 rare fragile sites, encompassing more than 100 Mb of DNA are currently listed in the human genome database. Common or constitutive fragile sites are chromosome regions, observed in metaphase chromosomes of all
0304-3835/$ - see front matter q 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.06.048
28
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
individuals tested; that are more susceptible to breakage, rearrangements and deletions than other sites of the genome. These sites appear in metaphase chromosomes as gaps or breaks triggered by treatment of cells with agents that delay replication, such as the DNA polymerase inhibitor aphidicolin (for review [1]). What these gaps represent at the molecular level is not known, but it was recently suggested that they represent single-stranded regions that have escaped notice by the ATR-CHK1 DNA damage checkpoint and have thus persisted into mitosis [2]. Shortly after the discovery of common fragile sites, it was observed that the cytogenetic locations of these fragile sites frequently coincided with the cytogenetic locations nonrandomly altered in specific cancers [3]. Thus, it was possible that the recombinogenicity of fragile regions predisposed them to chromosome rearrangements during cancer development. If a specific chromosome rearrangement is observed consistently in clonally expanded preneoplasias or neoplasias, it is likely that the genetic change resulting from the rearrangement has contributed to the clonal expansion. Thus, a number of investigators supposed that there must be genes at some common fragile regions, that when deleted, translocated or amplified due to fragile site instability, contribute to initiation or progression of cancer. Several laboratories thus cloned and characterized portions of the genomic regions in the vicinity of the most active fragile site, FRA3B [4–6] and sought to understand the relationship between fragility and specific DNA sequences. Even before common fragile sites were observed, a familial, balanced chromosome translocation involving chromosome regions 3p14.2 and 8q24, was reported in a family with early onset, bilateral and multifocal clear cell renal cancers [7]. Several fragile site laboratory groups were seeking to clone the t(3;8) translocation breakpoint, with the idea that a gene disrupted by the translocation break might be a tumor suppressor whose alteration had contributed to development of the familial kidney cancers [4–6]. The search culminated in 1996, in the discovery of the very large (1.6 Mb) FHIT gene that encompassed the fragile region, as well as the translocation breakpoint in intron 3 of the FHIT gene [8,9] (Fig. 1A). Cancer cell lines and primary cancers were shown to harbor deletions within one or both FHIT alleles [8–10], Ref. [1] for review) and large fractions of various
types of cancers, including upper gastrointestinal tract, colon, cervical, lung and breast cancers, expressed reduced or no Fhit protein (for review [11]). Although in the early years after discovery of the FHIT gene there was considerable resistance to acceptance of FHIT among the galaxy of Tumor Suppressor Genes [12,13] by early 2005 there were more than 500 reports concerning alteration of the FHIT gene or protein in cancers arising in almost every organ of the body. Nine years after its discovery, through contributions of many researchers from all over the globe, the role of FHIT as the archetypal fragile tumor suppressor gene is well established and these studies are serving as models for the isolation and characterization of genes at other common human fragile sites.
Fig. 1. Schematic of the FHIT and WWOX genes. The chromosome arm where each gene is located is shown on the left. The chromosome band where the genes are located is magnified and the genomic structure of each gene is shown on the right. The dark boxes represent protein coding exons and the open boxes non amino acid coding exons. Many of the deletions detected in tumor cell lines are centered on FHIT exon 5 and WWOX exon 8.
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
2. FHIT is down-regulated in most cancers Cancer cell lines and primary cancers exhibiting hemi or homozygous deletions with endpoints within the FHIT gene and reduced or absent Fhit expression have been reported (for review [1]). Reduced expression of Fhit in pre-invasive lesions is also observed. In esophageal cancer, Mori et al., reported that most of the in-situ lesions, 50% of severe and moderate dysplasias and 33% of mild dysplasias were Fhit negative [14]; in the study of Kitamura et al. [15] reduced Fhit expression was seen in 68% of in situ and 43.5% of esophageal dysplastic lesions. Hao et al. [16] found reduced Fhit expression only in a small fraction of adenomatous colon lesions, but reduced Fhit expression was associated with greater degree of dysplasia. In cervical cancer Connolly et al. [17] observed reduced or absent Fhit staining in 71% of invasive cancers and in 52% of high-grade intraepithelial lesions (HSILs) with invasive cancer. In w85% of bronchial dysphasia’s there was loss of Fhit expression [18]. In our study of DCIS (ductal carcinoma in situ), reduced Fhit expression was observed in: 70% of pure DCIS and 52% of DCIS adjacent-to-invasive tumor cases. 20% of pure DCIS cases exhibited individual glands of adjacent normal tissue with absence of expression [19]. LOH at 3p14.2 alleles has been found in pre-invasive bronchial lesions and the non-neoplastic airway cells of smokers and former smokers [11] for review). These clinical findings supported the proposal that FHIT inactivation occurs in early steps of carcinogenesis in many organs. The association between exposure to carcinogens and FHIT inactivation was first noted in lung cancers, which are among the cancers with the highest frequency of complete loss of Fhit expression [18], suggesting that alteration of the FHIT gene through damage to the associated fragile region by environmental carcinogens, contributes substantially to the human cancer burden. In support of this idea, FHIT inactivation was observed twice as often in lung tumors of smokers (75%) relative to non-smokers (39%) [18,20]. Exposure to asbestos and to gamma irradiation during the Chernobyl accident caused an increase in FHIT inactivation in lung cancer and preneoplastic bronchial lesions [20,21]. Smoking
29
history and alcohol abuse associated with higher frequency of loss of Fhit expression was also reported in esophageal cancer [14]. If FHIT is one of the first targets of carcinogens, the ability of the host to repair this initial damage or to eliminate cells carrying damage to the FHIT locus, may prevent clonal expansion. In support of this idea, loss of FHIT function is observed more frequently in cancers developing in individuals with constitutional alterations to genes involved in DNA repair, such as the BRCA1 gene and mismatch repair genes [22–24].
3. Epigenetic mechanisms contribute to loss of Fhit expression in cancers Epigenetic changes to chromatin, such as DNA methylation and modifications in histone tails regulate transcription of numerous tumor suppressor genes (for review [25]). DNA hypermethylation of promoter regions of tumor suppressor genes has been reported to be associated with decreased gene expression. FHIT promoter hypermethylation associated with loss of gene expression was reported for lung, breast and esophageal cancer [26–28]. It has also been reported that in non-small cell lung and bladder carcinomas FHIT promoter hypermethylation correlated with worse prognosis; in lung carcinomas, it was reported to be a sign of worse prognosis, even in patients with earlier stage of disease [28,29]. Hypermethylation of FHIT was associated with pack-years of cigarette smoking in human lung squamous cell carcinomas [30,31]. In a study of rodent tumors, most showed reduced Fhit expression and methylation in promoter, exon 1 and intron 1 regions [32]. In rat primary tumors induced by different carcinogens, different CpG sites were methylated, suggesting tissue or carcinogenspecific methylation patterns; in a follow-up study, it was determined that Fhit methylation patterns in benign and malignant tumors induced by two different carcinogens (MNU, DMBA) were tissue, not carcinogen, specific [33]. In human cancers we observed differential patterns of FHIT methylation in neoplastic vs. non neoplastic tissues; FHIT intron 1 methylation may serve as a marker of early steps in breast cancer development, suggesting that targeted methylationspecific PCR (MSP) amplification would be useful in following treatment or prevention protocols [28].
30
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
4. Fhit deficient mouse models to study cancer development, prevention and therapy Because the murine Fhit gene encompasses a common fragile site, Fra 14A2 and homozygous deletions of Fhit exons were observed in two mouse cancer lines (reviewed in [11]), the mouse model was chosen to study Fhit inactivation effects. FhitK/K mice, although being healthy and fertile, showed increased susceptibility to spontaneous and carcinogen-induced tumors [34]. All Fhit C/K mice developed several forestomach and sebaceous gland tumors after six doses of intragastric N-nitrosomethylbenzylamine (NMBA) while only 25% of wildtype mice developed tumors. The tumors were a mixture of benign, in-situ and invasive lesions. The NMBAinduced tumor spectrum in FhitC/K mice was similar to a human syndrome called Muir-Torre syndrome, a variant of hereditary non-polyposis colorectal cancer, which is often caused by inactivation of a mismatch repair gene, usually MSH2. Homozygous deletions of FHIT were observed in half of the cancer cell lines with a mutant mismatch repair gene; among nine Msh2-negative human colon cancer cells eight were negative for Fhit, and Fhit loss was reported in 90% of BRCA1 and 2 associated cancers [22–24]. These correlations suggest that fragile genes are especially vulnerable to damage-induced alterations in cells with ‘caretaker’ gene deficiencies. After showing increased frequency of tumors in Fhit deficient mice, the next question was: can FHIT gene therapy prevent and reverse development of these tumors (for review, [35]). Esophageal cancer cell lines transfected with Adeno-FHIT virus exhibited reduced proliferation and less tumorigenic behaviour when injected into immunodeficient mice. The growth of xenografts of poorly differentiated pancreatic cancer cell lines in nude mice was suppressed by Adeno-FHIT and AAV-FHIT (adenoassociated-virus-containing FHIT infection). AdenoFHIT infection was also associated with increased survival of mice with disseminated pancreatic cancer. AAV-FHIT or Adeno-FHIT treated Fhit knockout mice developed significantly fewer and smaller tumors following NMBA exposure, suggesting that local delivery of the FHIT gene to human premalignant lesions might be an effective preventive strategy. Since FHIT is one of the first genes damaged in
response to carcinogen exposure, for example in normal and metaplastic respiratory epithelium of smokers, and is frequently inactivated in tumors of repair deficient individuals, FHIT gene therapy offers promise for treatment and prevention of various types of cancers.
5. Fhit protein function Despite strong evidence of Fhit tumor suppressor function, our knowledge of specific Fhit signal pathways and mechanisms involved in the suppressor activity is limited. It is well-known that overexpression of Fhit results in apoptosis in Fhit deficient cancer cells (for review [35]), and recent studies have demonstrated a role for Fhit in the responses to genotoxic damage induced by UVC light, mitomycin C, camptothecin and ionizing radiation [36,37]; w10-fold more colonies of Fhit-deficient cells survived exposure to high UVC doses (Fig. 2). After mitomycin C treatment w6-fold and after UVC treatment 3.5-fold more Fhit-positive human cancer cells than Fhit-negative cells had died and UVC surviving FhitK/K cells showed more than 5-fold increased mutation frequency. The histidine triad motif –HxHxH– of Fhit is conserved in all species. Fhit protein is homologous to S. pombe Aph1 enzyme, which has diadenosine triphosphate hydrolase activity, an enzymatic function that is conserved from yeast to human [1]. However, it has been shown that ‘hydrolase dead’ Fhit with the central histidine mutated to asparagine (H96N) suppresses tumorigenity as well as wild type protein. Structural studies show that binding of the Fhit dimer with two molecules of diadenosine triphoshate, results in highly phosphorylated surfaces, which may have signaling activity [38]. Very recently it was discovered that Fhit can be phosphorylated by Src and Src family kinases in vivo and in vitro at amino acid Y114 [39]. Because Fhit is a dimer, there are three forms of Fhit, unphosphorylated, monophosphorylated, with only one monomer of the dimer phosphorylated, and the diphosphorylated form. The phosphorylated Fhit monomer is upshifted on PAGE gels, so that phosphorylation is easily followed on immunoblots. Phosphorylation of endogenous Fhit is not observed in cells in tissue culture, thus the consequences of phosphorylation of the Y114
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Fig. 2. FhitK/K and C/C mouse kidney cells were seeded, exposed to 60 J/m2 UV, collected, counted and re-seeded. After 18 days the plates were fixed and stained with Giemsa, and colonies counted. FhitK/K cells formed w90 colonies and FhitC/C cells formed w10 colonies. B. MKN-74 E4 and MKN-74 A66 human gastric cancer cells were seeded, exposed to 60 J/m2 UV, collected, counted, re-seeded and allowed to form colonies. After 18 days plates were fixed and stained with Giemsa. The colony numbers for the E4 (Fhit negative cells) were about ten times greater than for A66 (Fhit positive) cells: E4 w300 colonies. A66 w30 colonies. This figure was first published as Fig. 3b in Ref. [36] (British J Cancer).
tyrosine residue has been difficult to examine in vitro. Mutant Fhit constructs, with Y114 and nearby amino acids altered, have been prepared in plasmids as well as Adenovirus vectors, in order to study the biological function of Fhit, with and without phosphorylation at Y114.
6. The WWOX tumor suppressor gene at the FRA16D fragile site 6.1. WWOX structure and expression in cancer The WWOX gene spans a genomic region of more than one million nucleotide base pairs located at 16q23.3-24.1, a chromosome region commonly
31
involved in LOH in many different types of cancer (Fig. 1B); it is composed of nine exons, encoding a cDNA of 1245 bp [40]. Studies of esophageal squamous cell carcinoma, non-small cell lung cancer and breast cancer showed a high LOH rate, low mutation rate and expression of aberrant transcripts of the WWOX gene (reviewed in [41]). Northern blot and RT-PCR analyses from breast and ovarian cancer cells revealed the presence of transcripts of smaller size, representing abnormally spliced versions of WWOX [40,43]. WWOX mRNA expression is reduced in a series of human breast tumors and breast cancer cell lines [41–45]. In the first study of Wwox expression in clinical cancer tissues, coordinate absence or reduction of Fhit and Wwox expression in 55 and 63% of invasive breast tumors was reported (examples in Fig. 3) [46]. Reduced Wwox staining was found more frequently in ER (K) or scanty positive tumors. In adjacent normal tissue, reduced Wwox staining was observed in 32.9% of cases. Of all cases that showed Wwox staining in less than 10% of normal breast tissue were postmenopausal or exposed to neoadjuvant chemotherapy, implying that Wwox expression may be associated with the level of steroid hormone expression and can be effected by chemotherapeutic drugs. Nunez et al. [47] also reported frequent loss of Wwox expression in association with decreased ER expression in breast cancer. In our study of DCIS, reduced Fhit and Wwox expression were also observed in: (a) 70 and 68% of pure DCIS; (b) 52 and 55% of DCIS adjacent-to-invasive tumor cases; (c) some individual glands of adjacent normal tissue of 20 and 50% of pure DCIS cases [19]. Reduced Wwox expression in adjacent normal tissue was observed in 30% of DCIS adjacent-to-invasive cases. Reduced Fhit and Wwox expression was observed in 61% of adjoining invasive tumors. In all normal, pure DCIS and DCIS adjacent-to-invasive lesions, Fhit and Wwox expression were positively associated. Fhit and Wwox were more frequently reduced in high-grade lesions in the DCIS adjacentto-invasive group. In summary, Fhit and Wwox were lost coordinately in in-situ breast cancer, losses that may contribute to the high grade DCIS-invasive tumor pathway. Frequent loss and downregulation of Wwox expression was reported in human hepatocellular carcinoma cell lines; in primary gastric cancers,
32
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Fig. 3. Immunohistochemical staining of Fhit and Wwox. (A) An example of strong Fhit staining intensity in tumor and normal luminal epithelial cells (between arrows). (B) Loss of Fhit staining in tumor cells. (C) Similarly reduced Wwox staining intensity in adjacent normal luminal epithelial cells and tumor tissue. (D) Reduced Wwox staining intensity of tumor cells (adjacent normal tissue is between arrows). (E) Reduced Wwox staining intensity in normal luminal epithelial cells. (F) Intense cytoplasmic staining of Wwox in !10% of neoplastic cells. Original magnification !100 (B,F); !200 (C–E); !400 (A). (The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinomas, Gulnur Guler, CANCER, 100(8): 1605–14, Wiley–Liss, Inc.).
LOH in WWOX locus was observed in 31% and loss of Wwox protein expression in 65% of cases and in the same study a high correlation between Fhit and Wwox expression levels was found (reviewed in [41]).
7. Epigenetic mechanisms regulate WWOX expression Expression of many tumor suppressor genes is down-regulated in cancer by epigenetic mechanisms;
Fig. 4. Wwox regulatory region. The highest percentage of CpG sites starts from K183 to C177 relative to the transcription initiation site. 3 CpGs in the promoter region correspond to Sp1 binding sites.
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
CpG rich areas (islands) of gene promoter regions are frequently methylated in cancer (for review, [25]), and histone modifications, such as deacetylation and lysine methylation, change the chromatin structure in the promoter region of tumor suppressor genes, so that transcription factors cannot bind and transcription is blocked. Significantly, modifications are reversible by demethylating agents and HDAC inhibitors and are potential targets for cancer therapy. We inspected an interval including 1 kb upstream and downstream from the WWOX transcription initiation site for in silico CpG island identification [48]. A region 406 bp upstream of the WWOX transcription initiation site, exon 1 and part of intron 1 (the first 125 bp) exhibited 66% CG residues, the region from K183 to C177 has the highest percentage (w70%) of CpG sites (Fig. 4). Ishii et al., reported a possible role of epigenetic mechanisms in WWOX transcriptional regulation [49]. Specifically treatment of K562 leukemia cells with 5-Aza-2-deoxycytidine (demethylating agent) and depsipeptide (HDAC inhibitor) increased WWOX mRNA expression. Kuroki et al. [50] studied WWOX promoter methylation status in pancreatic cancer and reported methylation in the region K148 to K37 respective to the transcription initiation site (C1) in pancreatic adenocarcinomas; WWOX promoter methylation status was associated with its expression and after treatment with 5-Aza-deoxycytidine in Hs766T cells, Wwox was re-expressed. We have been interested in mechanisms that control WWOX fragile gene expression and inactivation in human cancers and have assessed lung, breast and bladder cancers for control of expression by methylation of 5 0 CpG islands. Frozen lung and breast tumor tissue, adjacent non-neoplastic and normal tissue samples were used in this study [48]. Wwox expression was reduced in breast and lung cancers in association with hypermethylation. Differential patterns of WWOX methylation were observed in neoplastic vs adjacent non-neoplastic tissues, suggesting that targeted MSP amplification could be useful in following treatment or prevention protocols. WWOX promoter MSP differentiates DNA of lung cancer from DNA of adjacent lung tissue. WWOX and FHIT promoter methylation is detected in tissue adjacent to breast cancer and WWOX exon 1 MSP distinguishes breast cancer DNA from DNA of
33
adjacent and normal tissue. Differential methylation in cancerous vs adjacent tissues suggests that WWOX hypermethylation analyses could enrich a panel of DNA methylation markers.
8. Wwox protein structure and function Bednarek et al., described the WWOX gene at 16q23.3-24.1 and observed that Wwox protein contains two WW domains in the NH2 terminus and a SDR (short chain dehydrogenase/reductase) domain [40]. WW domains are globular domains consisting of w40 amino acids, of which two tryptophans and an invariant proline are highly conserved [51]. Like SH3 domains, the WW domains are characterized by interactions with proline-containing ligands and mediate protein–protein interactions. WW domains can be grouped into four classes according to their ligand binding preferences and recently it was suggested that WW domain interactions can be modulated by tyrosine phosphorylation [52]. The short-chain dehydrogenase reductase (SDR) family includes steroid dehydrogenase enzymes and more than 60 other proteins from human, mammalian, insect, and bacterial sources. Most family members contain tyrosine and lysine of the catalytic triad in a YxxxK sequence. X-ray crystal structures of 13 members of the family showed that when the alphacarbon backbone of the cofactor binding domains of the structures are superimposed, the conserved residues are at the core of the structure and in the cofactor binding domain, but not in the substrate binding pocket [53]. The SDR domain in Wwox protein may play an important role in sex hormone regulated cancers such as breast, ovarian and prostate cancer. Chang et al. [54] recently reported that Wwox is upregulated in COS7 fibroblasts and DU145 prostate cancer cells after 17-b-estradiol treatment. WWOX spans the common fragile site FRA16D and it has been reported that FRA16D and the corresponding gene, WWOX, are highly conserved in the mouse at Fra8E1 (for review [41]). In studies seeking the function of Wwox, Ludes-Meyers et al. [55] showed that Wwox contains a Group I WW domain that binds known cellular proteins containing the specific ligand PPXY.
34
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
Aqeilan et al., [56] demonstrated a physical interaction between the first WW domain of Wwox and p73, a p53 homolog. The PPPY motif in the WW1 domain can be phosphorylated by Src-family tyrosine kinases, perhaps indicating that Src directly phosphorylates this residue. Overexpression of Wwox caused redistribution of p73 from the nucleus to the cytoplasm and suppression of p73 mediated interactions. Wwox also binds to the PPPY motif of AP2g via its first WW domain [57]. Ap2g encodes a transcription factor and is frequently up-regulated in breast carcinomas. Wwox overexpression triggered redistribution of nuclear Ap2g to the cytoplasm, hence suppressing its transactivating function. Interestingly, point mutation in the terminal tyrosine of the PPPY motif abrogates Wwox-p73 and -Ap2g interactions, emphasizing the principal of sequence-specific protein–protein interactions as a determinant of precise biological outputs. Thus Wwox appears to act as a modulator of transcription by binding these proteins. Our limited information about Wwox function suggests that, like Fhit, Wwox may participate in apoptotic pathways, though details of the Wwox signal pathways are not yet known.
9. Conclusions The apparent coordinate inactivation of the two fragile genes, FHIT and WWOX, in breast cancer, at the protein expression and promoter methylation level, [19,33], suggests the possibility of coordinate carcinogen damage to multiple fragile sites in some types of cancer. As we learn more about the fragile sites and the functions of genes encompassing them, we will gain further insight into their roles in initiation and progression of carcinogenesis.
Acknowledgements This work was supported by DOD predoctoral grant BC043090 (D.I), support from Hacettape University and the National Cancer Institute, USA (G.G), T32-HLO7780 (K.A.M) and the NCI grants CA77738 and CA56036.
References [1] K. Huebner, C.M. Croce, FRA3B and other common fragile sites: the weakest links, Nat. Rev. Cancer 1 (2001) 214–221. [2] A.M. Casper, P. Nghiem, M. Arlt, T.W. Glover, ATR regulates fragile site stability, Cell 111 (2002) 779–789. [3] J. Yunis, A.L. Soreng, Constitutive fragile sites and cancer, Science 226 (1984) 1199–1204. [4] F.L. Boldog, B. Waggoner, T.W. Glover, I. Chumakov, D. Le Paslier, D. Cohen, et al., Integrated YAC contig containing the 3p14.2 hereditary renal carcinoma 3;8 translocation breakpoint and the fragile site FRA3B, Genes Chromosomes Cancer 11 (1994) 216–221. [5] C.M. Wilke, S.W. Guo, B.K. Hall, F. Boldog, R.M. Gemmill, S. Chandrasekharappa, et al., Multicolor FISH mapping of YAC clones in 3p14 and identification of a YAC spanning both FRA3B and the t(3;8) associated with hereditary renal cell carcinoma, Genomics 22 (1994) 319–326. [6] W. Paradee, C.M. Wilke, L. Wang, R. Shridhar, C.M. Mullins, A. Hoge, et al., A 350-kb cosmid contig in 3p14.2 that crosses the t(3;8) hereditary renal cell carcinoma translocation breakpoint and 17 aphidicolin-induced FRA3B breakpoints, Genomics 35 (1996) 87–93. [7] A.J. Cohen, F.P. Li, S. Berg, D.J. Marchetto, S. Tsai, S.C. Jacobs, R.S. Brown, Hereditary renal-cell carcinoma associated with a chromosomal translocation, N. Engl. J. Med. 301 (1979) 592–595. [8] M. Ohta, H. Inoue, M.G. Cotticelli, K. Kastury, R. Baffa, J. Palazzo, et al., The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers, Cell 84 (1996) 587–597. [9] D. Zimonjic, T. Druck, M. Ohta, K. Kastury, C.M. Croce, N.C. Popescu, K. Huebner, Positions of chromosome 3p14.2 fragile sites (FRA3B) within the FHIT gene, Cancer Res. 57 (1997) 1166–1170. [10] G. Sozzi, S. Tornielli, E. Tagliabue, L. Sard, F. Pezzella, U. Pastorino, et al., Absence of Fhit protein in primary lung tumors and cell lines with FHIT gene abnormalities, Cancer Res. 57 (1997) 5207–5212. [11] K. Huebner, C.M. Croce, Cancer and the FRA3B/FHIT fragile locus: it’s a HIT, Br. J. Cancer 10 (2003) 1501–1506. [12] S. Thiagalingam, N.A. Lisitsyn, M. Hamaguchi, M.H. Wigler, J.K. Willson, S.D. Markowitz, et al., Evaluation of the FHIT gene in colorectal cancers, Cancer Res. 56 (1996) 2936–2939. [13] M.M. Le Beau, H. Drabkin, T.W. Glover, R. Gemmill, F.V. Rassool, T.W. McKeithan, D.I. Smith, An FHIT tumor suppressor gene? Genes Chromosomes Cancer 21 (1998) 281–289. [14] M. Mori, K. Mimori, T. Shiraishi, H. Alder, H. Inoue, Y. Tanaka, et al., Altered expression of Fhit in carcinoma and precarcinomatous lesions of the esophagus, Cancer Res. 60 (2000) 1177–1182. [15] A. Kitamura, K. Yashima, E. Okamoto, H. Andachi, A. Hosoda, Y. Kishimoto, et al., Reduced Fhit expression
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25] [26]
[27]
[28]
[29]
occurs in the early stage of esophageal tumorigenesis: no correlation with p53 expression and apoptosis, Oncology 61 (2001) 205–211. X.P. Hao, J.E. Willis, T.G. Pretlow, J.S. Rao, G.T. MacLennan, I.C. Talbot, T.P. Pretlow, Loss of fragile histidine triad expression in colorectal carcinomas and premalignant lesions, Cancer Res. 60 (2000) 18–21. D.C. Connolly, D.L. Greenspan, R. Wu, X. Ren, R.L. Dunn, K.V. Shah, et al., Loss of Fhit expression in invasive cervical carcinomas and intraepithelial lesions associated with invasive disease, Clin. Cancer Res. 6 (2000) 3505–3510. G. Sozzi, U. Pastorino, L. Moiraghi, E. Tagliabue, F. Pezzella, C. Ghirelli, et al., Loss of FHIT function in lung cancer and preinvasive bronchial lesions, Cancer Res. 58 (1998) 5032– 5037. G. Guler, A. Uner, N. Gu¨ler, S.Y. Han, D. Iliopoulos, P. McCue, K. Huebner, Concordant loss of fragile gene expression early in breast cancer development, Pathol. Int. 8 (2005) 471–478. L. Pylkkanen, H. Wolff, T. Stjernvall, P. Tuominen, T. Sioris, A. Karjalainen, et al., Reduced Fhit protein expression and loss of heterozygosity at FHIT gene in tumours from smoking and asbestos-exposed lung cancer patients, Int. J. Oncol. 20 (2002) 285–290. V. Chizhikov, S. Chikina, A. Gasparian, I. Zborovskaya, E. Steshina, G. Ungiadze, et al., Molecular follow-up of preneoplastic lesions in bronchial epithelium of former Chernobyl clean-up workers, Oncogene 21 (2002) 2398–2405. S. Ingvarsson, B.A. Agnarsson, B.I. Sigbjornsdottir, J. Kononen, O.P. Kallioniemi, R.B. Barkardottir, et al., Reduced Fhit expression in sporadic and BRCA2-linked breast carcinomas, Cancer Res. 59 (1999) 2682–2689. M. Mori, K. Mimori, T. Masuda, K. Yoshinaga, K. Yamashita, A. Matsuyama, H. Inoue, Absence of Msh2 protein expression is associated with alteration in the FHIT locus and Fhit protein expression in colorectal carcinoma, Cancer Res. 61 (2001) 7379–7382. B.C. Turner, M. Ottey, D.B. Zimonjic, M. Potoczek, W.W. Hauck, E. Pequignot, et al., The fragile histidine triad/common chromosome fragile site 3B locus and repairdeficient cancers, Cancer Res. 62 (2002) 4054–4060. C.L. Hsieh, P.A. Jones, Meddling with methylation, Nat. Cell Biol. 5 (2003) 502–504. H. Tanaka, Y. Shimada, H. Harada, M. Shinoda, S. Hatooka, M. Imamura, K. Ishizaki, Methylation of the 5 0 CpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas, Cancer Res. 58 (1998) 3429–3434. S. Zochbauer-Muller, K.M. Fong, A. Maitra, S. lam, J. Goradts, R. Ashfuq, et al., 5 0 CpG island methylation of the FHIT gene is correlated with loss of gene expression in lung and breast cancer, Cancer Res. 61 (2001) 3581–3585. R. Maruyama, K. Sugio, I. Yoshino, Y. Maehara, A.F. Gazdar, Hypermethylation of FHIT as a prognostic marker in nonsmall cell lung carcinoma, Cancer 100 (2004) 1472–1477. R. Maruyama, S. Toyooka, K.O. Toyooka, K. Harada, A.K. Virmani, S. Zochbauer-Muller, et al., Aberrant promoter
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
35
methylation profile of bladder cancer and its relationship to clinicopathological features, Cancer Res. 61 (2001) 8659– 8663. J.S. Kim, H. Kim, Y.M. Shim, J. Han, J. Park, D.H. Kim, Aberrant methylation of the FHIT gene in chronic smokers with early stage squamous cell carcinoma of the lung, Carcinogenesis 25 (2004) 2165–2171. H. Kim, Y.M. Kwon, J.S. Kim, H. Lee, J.H. Park, Y.M. Shim, et al., Tumor-specific methylation in bronchial lavage for the early detection of non-small-cell lung cancer, J. Clin. Oncol. 22 (2004) 2363–2370. S.Y. Han, D. Iliopoulos, T. Druck, G. Guler, C.J. Grubbs, M. Pereira, et al., CpG methylation in the Fhit regulatory region: relation to Fhit expression in murine tumors, Oncogene 23 (2004) 3990–3998. G. Guler, D. Iliopoulos, S.Y. Han, L.Y. Fong, R.A. Lubet, C.J. Grubbs, K. Huebner, Hypermethylation patterns in the Fhit regulatory region are tissue specific, Mol. Carcinog. 3 (2005) 175–181. N. Zanesi, V. Fidanza, L.Y. Fong, R. Mancini, T. Druck, M. Valtieri, et al., The tumor spectrum in FHIT-deficient mice, Proc. Natl. Acad. Sci. USA 98 (2001) 10250–10255. H. Ishii, K.R. Dumon, A. Vecchione, L.Y. Fong, R. Baffa, K. Huebner, C.M. Croce, Potential cancer therapy with the fragile histidine triad gene: review of the preclinical studies, J. Am. Med. Assoc. 19 (2001) 2441–2449. M. Ottey, S.Y. Han, T. Druck, B.L. Barnoski, K.A. McCorkell, C.M. Croce, et al., Fhit-deficient normal and cancer cells are mitomycin C and UVC resistant, Br. J. Cancer 9 (2004) 1669–1677. B. Hu, S.Y. Han, X. Wang, M. Ottey, M.B. Potoczek, A. Dicker, et al., Involvement of the Fhit gene in the ionizing radiation-activated ATR/CHK1 pathway, J. Cell Physiol. 202 (2005) 518–523. H.C. Pace, P.N. Garrison, A.K. Robinson, L.D. Barnes, A. Draganescu, A. Rosler, et al., Genetic, biochemical, and crystallographic characterization of Fhit-substrate complexes as the active signaling form of Fhit, Proc. Natl. Acad. Sci. USA 95 (1998) 5484–5489. Y. Pekarsky, P.N. Garrison, A. Palamarchuk, N. Zanesi, R.I. Aqeilan, K. Huebner, et al., Fhit is a physiological target of the protein kinase Src, Proc. Natl. Acad. Sci. USA 101 (2004) 3775–3779. A.K. Bednarek, K.J. Laflin, R.L. Daniel, Q. Liao, K.A. Hawkins, C.M. Aldaz, WWOX, a novel WW domaincontaining protein mapping to human chromosome 16q23.324.1, a region frequently affected in breast cancer, Cancer Res. 60 (2000) 2140–2145. J.H. Ludes-Meyers, A.K. Bednarek, N.C. Popescu, M. Bedford, C.M. Aldaz, WWOX, the common chromosomal fragile site, FRA16D, cancer gene, Cytogenet. Genome Res. 100 (2003) 101–110. A.K. Bednarek, C.L. Keck-Waggoner, R.L. Daniel, K. Laflin, P. Bergsagel, K. Kiguchi, et al., FRA16D gene, behaves as a suppressor of tumor growth, Cancer Res. 61 (2001) 8068–8073. A.J. Paige, K.J. Taylor, A. Stewart, J.G. Sgouros, H. Gabra, G.C. Sellar, et al., A 700-kb physical map of a region of
36
[44]
[45]
[46]
[47]
[48]
[49]
D. Iliopoulos et al. / Cancer Letters 232 (2006) 27–36 16q23.2 homozygously deleted in multiple cancers and spanning the common fragile site FRA16D, Cancer Res. 60 (2000) 1690–1697. M. Mangelsdorf, K. Ried, E. Woollatt, S. Dayan, H. Eyre, M. Finnis, et al., Chromosomal fragile site FRA16D and DNA instability in cancer, Cancer Res. 60 (2000) 1683–1689. A.K. Bednarek, C.L. Keck-Waggoner, R.L. Daniel, K.J. Laflin, P.L. Bergsagel, K. Kiguchi, et al., FRA16D gene, behaves as a suppressor of tumor growth, Cancer Res. 61 (2001) 8068–8073. G. Guler, A. Uner, N. Guler, S.Y. Han, D. Iliopoulos, W.W. Hauck, et al., The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinoma, Cancer 8 (2004) 1605–1614. M.I. Nunez, J. Ludes-Meyers, M.C. Abba, H. Kil, N. Abbey, R.E. Page, et al., Frequent loss of WWOX expression in breast cancer: correlation with estrogen receptor status, Breast Cancer Res. Treat. 7 (2005) 32–1614. D. Iliopoulos, G. Guler, S.Y. Han, D. Johnston, T. Druck, K.A. McCorkell, et al., Fragile genes as biomarkers: epigenetic control of WWOX and FHIT in lung, breast and bladder cancer, Oncogene 9 (2005) 1625–1633. H. Ishii, K. Mimor, T. Inageta, Y. Murakumo, A. Vecchione, M. Mori, Y. Furukawa, Components of DNA damage checkpoint pathway regulate UV exposuredependent alterations of gene expression of FHIT and WWOX at chromosome fragile sites, Mol. Cancer Res. 3 (2005) 130–138.
[50] T. Kuroki, S. Yendamuri, F. Trapasso, A. Matsuyama, R.I. Aqeilan, H. Alder, et al., The tumor suppressor gene WWOX at FRA16D is involved in pancreatic carcinogenesis, Clin. Cancer Res. 10 (2004) 2459–2465. [51] M.J. Macias, WW and SH3 domains, two different scaffolds to recognize proline-rich ligands, FEBS Lett. 513 (2002) 30–37. [52] J.L. Ilsley, M. Sudol, S.J. Winder, The WW domain: linking cell signalling to the membrane cytoskeleton, Cell signaling 14 (2002) 183–189. [53] W.L. Duax, D. Ghosh, V. Pletnev, Steroid dehydrogenase structures, mechanism of action, and disease, Vitam Horm. 58 (2000) 121–148. [54] N.S. Chang, L. Schultz, L.J. Hsu, J. Lewis, M. Su, C.I. Sze, 17beta-Estradiol up-regulates and activates WOX1/WWOXv1 and WOX2/WWOXv2 in vitro: potential role in cancerous progression of breast and prostate to a premetastatic state in vivo, Oncogene 24 (2005) 714–723. [55] J.H. Ludes-Meyers, H. Kil, A.K. Bednarek, J. Drake, M. Bedford, C.M. Aldaz, W.W.O.X. binds, the specific proline-rich ligand PPXY: identification of candidate interacting proteins, Oncogene 23 (2004) 5049–5055. [56] R.I. Aqeilan, Y. Pekarsky, J.J. Herrero, A. Palamarchuk, J. Letofsky, T. Druck, Functional association between Wwox tumor suppressor protein and p73, a p53 homolog, Proc. Natl. Acad. Sci. USA 101 (2004) 4401–4406. [57] R.I. Aqeilan, A. Palamarchuk, R.J. Weigel, J.J. Herrero, Y. Pekarsky, C.M. Croce, Physical and functional interactions between the Wwox tumor suppressor protein and the AP2gamma transcription factor, Cancer Res. 64 (2004) 8256–8266.