THE ROLE OF THE WWOX TUMOR SUPPRESSOR
IN BREAST AND LUNG CANCER
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of the Ohio State University
By
Dimitrios Iliopoulos, B.Sc
*****
The Ohio State University
2006
Dissertation Committee
Dr Kay Huebner, Adviser Approved by Dr Tim Huang Dr Danilo Perrotti Dr Helen Pace Adviser Dr Samir Acharya Intergrated Biomedical Science Graduate Program
ABSTRACT
The WWOX gene spans a genomic region of more than half million nucleotide base pairs located at 16q23.3-24.1, a chromosome region commonly involved in LOH in many different types of cancer. Wwox is frequently down-regulated in breast and lung cancers due to DNA hypermethylation in its promoter region. We observed differential patterns of WWOX methylation in neoplastic vs. adjacent non neoplastic tissues, suggesting that targeted WWOX methylation specific amplification could be useful in following treatment or prevention protocols, and WWOX methylation analyses could enrich a panel of DNA methylation markers. Restoration of Wwox expression in Wwox- negative breast and lung cancer-derived cells suppressed tumor growth in vitro and in vivo and induced apoptosis, confirming that WWOX is a tumor suppressor gene that is highly effective in gene therapy of breast and lung cancer xenografts, whether transduced by adenovirus or re-expressed through epigenetic therapy. The preclinical studies that we have performed showed the therapeutic potential of restoration of tumor suppressor expression through epigenetic modulation and the promise of reexpressed tumor suppressors such as WWOX as markers and effectors of the responses. Although many
ii patients benefit from tamoxifen treatment, half of breast tumors that recur after therapy are resistant to tamoxifen. Understanding mechanisms of tamoxifen resistance could lead to characterization of protein markers for identification of nonresponsive cancers, as well as tumors that are acquiring resistance, before emergence of more aggressive cancer cells.
We have found that Wwox mediates the tamoxifen response through regulation of protein kinase A and ErbB2 signaling pathways and high Wwox expression level predicted tamoxifen sensitivity in a cohort of breast cancer cases. The results imply that epigenetic reactivation of Wwox could sensitize tamoxifen resistant cells, suggesting possible epigenetic and hormonal combination therapy in patients with acquired tamoxifen resistance.
iii
Dedicated to my parents, Charalampos and Eleni
and my brother, Achilleas
iv ACKNOWLEDGMENTS
I would like to thank my adviser, Dr Kay Huebner, for her intellectual support, encouragement and enthusiasm, which made this thesis possible. I could not have worked with a greater mentor for these past few years. Also I would like to thank my committee members Drs Tim H. Huang, Danilo Perrotti, Helen Pace and Samir Acharya for their comments and suggestions.
I am grateful to the current and former members of our lab for their suggestions regarding experimental design and troubleshooting including Rebekah Qin, Jin Sun,
Susho Semba and Kelly Ann McCorkell. I would like to thank especially Teresa Druck for her continuous help, it was a pleasure for me to work with her and having scientific discussions about designing experiments, writing papers or grants. Also I would like to acknowledge Dr Shuang-Yin Han who helped me to improve technically in the first months of my doctorate.
In addition I would like to thank Dr Gulnur Guler for her collaboration and help in the breast cancer projects. She influenced me positively to focus on breast cancer, which is an exciting field of cancer study. Also I would like to acknowledge our collaborators Juan Palazzo, Peter A. McCue, Raffaele Baffa, Danika Johnston, Joshua
Cantor, Meiyun Fan, Kenneth Nephew, Ronald Weigel and especially Dr Carlo Croce lab
v members including Gianpiero Di Leva, Muller Fabbri, Fabio Petrocca, Flavia Pichiorri,
Rosa Visone, Nichola Zanesi and Alessandra Drusco.
Finally I would like to thank my family for their support all these years, which was very important to me.
vi VITA
November 24, 1978……………………………...Born- Larisa, Greece
1997 – 2001 ……………………………………..B.Sc Biology, Aristotle University
2002 – present …………………………………...Graduate student in IBGP program at
the Ohio State University
PUBLICATIONS
1. Iliopoulos D, Volakakis N, Tsiga A, Rousso I and Voyiatzis N. Description and molecular analysis in a patient with 46,XY pure gonadal dysgenesis (Swyer syndrome). Annales de genetique 47: 185-190, 2004.
2. Iliopoulos D, Poulsides G, Peristeri V, Kouri G, Andreou A and Voyiatzis N. Double trisomy (48,XXY,+21) in monozygotic twins: case report and review of the literature. Annales de genetique 47: 95-98, 2004.
3. Han SY*, Iliopoulos D*, Druck T, Guler G, Grubbs CJ, Pereira M, Zhang Z, You M, Lubet RA, Fong LY, Huebner K. CpG methylation in the Fhit regulatory region: relation to Fhit expression in murine tumors. Oncogene 23: 3990- 8, 2004. *contributed equally.
4. Guler G, Uner A, Guler U, Han SY, Iliopoulos D, Hauck W.W, McCue P, Huebner K. The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinoma: Correlations with clinical features. Cancer. 100 (8) 1605-14, 2004.
5. Iliopoulos D, Guler G, Han SY, Johnston D, Druck T, McCorkell KA, Palazzo J, McCue P, Baffa R, Huebner K. Fragile genes as potential biomarkers: Epigenetic control of FHIT and WWOX in lung, breast and bladder cancer. Oncogene 24: 1625-33, 2005.
vii 6. Guler G*, Iliopoulos D*, Han SY, Fong LY, Lubet RA, Grubbs CJ and Huebner K. Hypermethylation patterns in the Fhit regulatory region are tissue specific. Molecular Carcinogenesis 43:175-81, 2005. *contributed equally.
7. Guler G, Uner A, Guler N, Han SY, Iliopoulos D, McCue PA, Huebner K. Concodant loss of fragile gene expression early in breast cancer development. Pathology International 55:471-8, 2005.
8. Fabbri M*, Iliopoulos D*, Trapasso F, Aqeilan RI, Cimmino A, Zanesi N, Yendamuri S, Han SY, Amadori D, Huebner K and Croce CM. WWOX gene therapy prevents lung cancer growth in vitro and in vivo. Proc Natl Acad Sci USA 102: 15611-6, 2005. *contributed equally.
9. Iliopoulos D, Vassiliou G, Sekerli E, Sidiropoulou V, Tsiga A, Dimopoulou D, Voyiatzis N. Long survival in a 69,XXX triploid infant in Greece. Am J. Med Genet, 1: 92-3, 2005.
10. Iliopoulos D, Guler G, Han SY, Druck T, McCorkell KA and Huebner K. Role of FHIT and WWOX fragile genes in cancer, Cancer Lett 232: 27-36, 2006.
11. Guler G, Iliopoulos D, Han SY, Druck T, McCorkell KA and Huebner K. Fragile genes at common fragile sites: roles in carcinogenesis, book chapter, 2006.
12. Semba S, Trapasso F, Fabbri M, McCorkell KA, Volinia S, Druck T, Iliopoulos D, Ishii H, Barnes LD, Croce CM and Huebner K. Fhit modulation of Akt-Survivin pathway in lung cancer cells: FhitY114 is essential. Oncogene (epub Jan, 2006).
13. Satra M, Gatselis N, Iliopoulos D, Zacharoulis D, Dalekos GN, Tsezou A. Real time quantification of human telomerase reverse transcriptase mRNA in liver tissues from patients with hepatocellular cancer and chronic viral hepatitis. J Viral Hepat (in print).
14. Pichiorri F, Trapasso F, Palombo T, Aqeilan RI, Drusco A, Blaser BW, Iliopoulos D, Caligiuri MA, Huebner K, Croce CM. Preclinical assessment of FHIT gene replacement therapy in human leukemia using a chimeric adenovirus, Ad5/F35. Clin Cancer Res (in print).
viii 15. Cantor JP*, Iliopoulos D*, Rao AS, Druck T, Semba S, Han SY, McCorkell KA, Lakshman TV, Collins JE, Wachsberger P, Friedberg JS, Huebner K. Epigenetic modulation of tumor suppressor expression in lung cancer xenografts suppresses tumorigenicity Int J Cancer (in print). *contributed equally
16. Qin HY, Iliopoulos D, Semba S, Fabbri M, Druck T, Volinia S, Croce CM, Morrison CD, Klein RD, and Huebner K. Role of WWOX tumor suppressor gene in prostate cancer Cancer Res (in print).
17. Semba S, Han SY, Qin H, McCorkell KA, Iliopoulos D, Pekarsky Y, Druck T, Trapasso F, Croce CM and Huebner K. Biological functions of mammalian Nit1, the counterpart of NitFhit Rosetta Stone protein: a possible tumor suppressor J Biol Chem (in print).
FIELDS OF STUDY
Major Field: Integrated Biomedical Graduate Program
ix TABLE OF CONTENTS
Page
Abstract……………………………………………………………………………...... ii
Dedication………………………………………………………………………………...iv
Acknowledgements………………………………………………………………………..v
Vita………………………………………………………………………………………vii
List of Tables…………………………………………………………………………...xvii
List of Figures………………………………………………………………………….xviii
Chapters
1. Introduction………………………………………………………………………. 1
1.1 Common fragile sites and their relation to cancer………….…..…………….1
1.2 FRA3B: the most active common fragile site…….…………………………..3
1.3 WWOX tumor suppressor gene structure and expression in cancer……..…..4
1.4 Epigenetic regulation of WWOX expression in cancer………………….…...7
1.5 WWOX protein structure and function in cancer…………...………………..9
2. Wwox is epigenetically regulated in cancer…………………………………..…11
2.1 Abstract…………………………………………………………………...….11
x 2.2 Introduction……………………………………………………………..……12
2.3 Materials and Methods…………………………………………………….…13
2.3.1 Patient samples………………………………………………….….13
2.3.2 In silico CpG island identification…………………………………14
2.3.3 Methylation specific PCR (MSP)………………………………… 14
2.3.4 Bisulfite treatment and sequencing…………………………...……16
2.3.5 Immunohistochemistry………………………………………….…16
2.4 Results………………………………………………………………….….…16
2.4.1 Wwox and Fhit expression…………...……………………………16
2.4.2 FHIT and WWOX CpG islands……………………………………21
2.4.3 WWOX and FHIT methylation status……………………………..21
2.4.4 Correlation of WWOX and FHIT methylation status and association
with clinicopathological features…………………………………..……22
2.4.5 Methylation patterns assessed by sequencing………..………...…24
2.4.6 WWOX and FHIT promoter methylation in cancers correlates with
reduced protein expression…………..……………………………....…..25
2.5 Discussion……………………………………………………………………28
xi
3. Adenoviral WWOX restoration suppresses tumorigenicity in vitro and in vivo...31
3.1 Abstract………………………………………………………………...…….31
3.2 Introduction…………………………………………………………..………32
3.3 Materials and methods……………………………………………………….33
3.3.1 Cell culture…………………………………………………………33
3.3.2 Recombinant Adenoviruses and In vitro Transduction……………34
3.3.3 Inducible WWOX transfectants……………………………………35
3.3.4 Western blot Analysis………………………………………...……35
3.3.5 Cell Growth and cell cycle kinetics……………………………..…35
3.3.6 In vivo studies……………………………………………...………36
3.3.7 Ex vivo studies…………………………………………………...…37
3.3.8 Statistical analysis……………………………………………….…37
3.4 Results……………………………………………………………………..…37 3.4.1 Wwox expression in parental and Ad-WWOX infected lung cancer
cells………………………………………………………….……37
3.4.2 Cell cycle kinetics of infected cells………………………………..39
3.4.3 Apoptotic pathways in Wwox-reexpressing cells…………….……39
3.4.4 Effects of conditional Wwox expression in H1299 cells………..…42
3.4.5 Tumorigenicity of Ad-WWOX-infected lung cancer cell lines……42
xii 3.4.6 Effect of drug-induced Wwox expression on tumorigenicity…...…45
3.4.7 Wwox expression in H1299/I+ explanted tumors…………………45
3.5 Discussion……………………………………………………………………47
4. Epigenetic restoration of WWOX suppresses tumorigenicity in vitro and in
vivo………………………………………………………………………………50
4.1 Abstract………………………………………………………………………50
4.2 Introduction………………………………………………………………..…51
4.3 Materials and methods…………………………………………………….…53
4.3.1 Cell culture and isolation of conditional H1299 Fhit expressing
clones………………………………………………………………….…53
4.3.2 In vitro induction and re-expression of Fhit……………………..…54
4.3.3 DNA isolation, bisulfite modification and methylation-specific
PCR……………………………………………………………………....54
4.3.4 Real-time PCR analysis……………………………………..……....55
4.3.5 Immunoblot Analysis……………….………………………………55
4.3.6 Animal Studies…………………………………………………..…55
4.3.7 Tumorigenicity studies of H1299 cells conditionally expressing
Fhit……………………………………………………………………….55
4.3.8 Tumorigenicity studies of H1299 cells: effects of epigenetic
therapy……………………………………………………………………56 xiii 4.3.9 Statistics……………………………………………………………57
4.4 Results………………………………………………………………….……58
4.4.1 In vitro Fhit induction in H1299D1 cells……………………….…58
4.4.2 Effect of Fhit induction on tumorigenesis…………………………58
4.4.3 Hypermethylation of suppressor gene promoters in H1299 cells….59
4.4.4 Fhit and Wwox protein re-expression in H1299 cells following AZA
and TSA treatment…………………………………………….…65
4.4.5 Effect of AZA and TSA treatment on tumorigenesis…………...…65
4.5 Discussion……………………………………………………………………71
5. The role of WWOX in hormone resistant breast cancer……………………..….76
5.1 Abstract…………………………………………………………………...….76
5.2 Introduction………………………………………………………………..…76
5.3 Materials and methods……………………………………………….………78
5.3.1 Antisera and drugs…………………………………………………78
5.3.2 Cell culture…………………………………………………………79
5.3.3 Establishment and characterization of the MCF7 cell line with
acquired tamoxifen resistance……………………………………79
5.3.4 RNA interference methods……………………………………...…81
5.3.5 Cell proliferation and death assays………………………………...81
xiv 5.3.6 Immunoblotting and immunoprecipitation……………………...…81
5.3.7 Immunofluorescence…………………………………………….…82
5.3.8 Luciferase assay……………………………………………………82
5.3.9 Protein Kinase A assay……………………………………….……83
5.3.10 Plasmid construction…………………………………………...…83
5.3.11 DNA methylation studies…………………………………………84
5.3.12 Adenoviral WWOX infections………………………………...…84
5.3.13 Immunohistochemical analysis…………………………..………84
5.3.14 Statistics…………………………………………………………..85
5.4 Results………………………………………………………………….……85
5.4.1Wwox down-regulation abrogates tamoxifen response……….……85
5.4.2 Wwox regulates ErbB2 and PKA signaling pathways………….…89
5.4.3 Restoration of tamoxifen effectiveness by Wwox up-regulation.…94
5.4.4 Epigenetic Wwox reactivation sensitizes tamoxifen resistant cells.96
5.4.5 Quantitative relationship between Wwox and tamoxifen
resistance………………………………………………………..103
5.4.6 Wwox protein levels and tamoxifen sensitivity in vivo…………..104
5.5 Discussion…………………………………………………………………..109
6. Conclusions and future directions………………………………………………115
6.1 WWOX methylation as a potential biomarker for lung and breast cancer...116
xv 6.2 WWOX gene therapy in lung and breast cancer………………………...…117
6.3 Wwox levels can predict tamoxifen sensitivity……………..………...... …119
6.4 Wwox as a regulator of major signaling pathways in breast cancer…….…120
Bibliography…………………………………………………………………………122
xvi LIST OF TABLES
Table Page
2.1 WWOX and FHIT primers for MSP and genomic bisulfite sequencing………15
2.2 Summary of FHIT and WWOX methylation status and expression…………...20
3.1 Tumor weight (g) ± SD in nude mice…………………………………………..46
xvii LIST OF FIGURES
Figure Page
1.1 Schematic of the FHIT and WWOX genes……………………..………………...…..5
1.2 Wwox regulatory region……………………….……………………………...……...8
2.1 Immunohistochemical detection of Fhit and Wwox expression in lung, breast and
bladder tissues…………..………………………………………….…..……………18
2.2 MSP amplification in WWOX and FHIT regulatory regions……….………………23
2.3 Methylation patterns in WWOX and FHIT 5’ regulatory regions……...………..….26
3.1 Expression of Wwox protein…………………………………………………….….37
3.2 Flow cytometry analysis of untreated, Ad-GFP and Ad-WWOX infected cells…....39
3.3 Effect of Wwox expression on cell growth in vitro………………..…………..……40
3.4 Inducible expression of Wwox in H1299/I cells……………………….……………42
3.5 Effect of Wwox expression on tumorigenicity of lung cancer cells………..……….43
3.6 Ex vivo analysis of H1299/I- and H1299/I+ cells………………………………...... 45
4.1 Effect of Fhit expression on tumorigenicity of lung cancer cells…………..……….59
4.2 FHIT and WWOX methylation status in cancer cells………………………………61
4.3 Effect of AZA and TSA treatment on FHIT methylation status and mRNA
expression…………………………………………………...………………………63
4.4 Fhit and Wwox protein re-expression in H1299 cells after AZA and TSA
treatment……………….……………………………………………………………66
xviii 4.5 In vivo epigenetic therapy of H1299 xenografts………………………..……….…..67
5.1 Wwox down-regulation abrogates tamoxifen sensitivity……………….…………..85
5.2 Wwox shRNA plasmid transfection abrogates tamoxifen response………..……….86
5.3 Identifying potential WWOX protein interactors………………………..………….89
5.4 Wwox mediation of ErbB2 and PKA activity…………………..…………………..90
5.5 Restoration of tamoxifen sensitivity by exogenous Wwox expression……….…….95
5.6 Wwox re-expression in BT-474 cells restores tamoxifen sensitivity……………….97
5.7 Adenoviral Wwox expression restores tamoxifen sensitivity in MCF7-OHTr cells..98
5.8 Wwox subcellular localization in MCF7-OHTr cells after WWOX plasmid
transfection……………………………………………………………...………...…99
5.9 Epigenetic Wwox re-activation sensitizes MCF7-OHTr cells to tamoxifen…….....100
5.10 Quantitative relationship between Wwox expression level and tamoxifen
sensitivity………………………………………….………………………………103
5.11 WWOX shRNA stable clones……………………………………………...……..105
5.12 Correlation of Wwox protein levels and tamoxifen sensitivity in breast cancer
tissues……………………………………………………………………..……...106
5.13 Model of Wwox-mediated tamoxifen resistance………………...……………….109
xix
CHAPTER 1
INTRODUCTION
1.1 Common fragile sites and their relation to cancer
Eighty common and twenty-four 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 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 (1, 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
1 regions, that when deleted, translocated or amplified due to fragile site instability, contribute to initiation or progression of cancer.
Common fragile sites show a consistent hierarchy in their relative frequency of expression with FRA3B being most frequently observed, followed by
FRA16D>FRA6E>FRA7G>FRAXB>others (4). Limited progress has been made in the identification of intrinsic (cis-acting) features of these common fragile site loci that specifically render them susceptible to breakage. In contrast, several trans-acting factors have been identified that regulate stability at these loci. Sequencing has revealed that common fragile site regions are predominantly characterized by the presence of various
AT-rich DNA sequences some of which comprise simple tandem repeats. In some cases, these sequences manifest as regions of high flexibility and low stability (5). It has been proposed that these AT-rich flexible sequences could be responsible for DNA instability and that the resultant fragility observed at common fragile sites could have a molecular basis similar to that of some rare fragile sites (6).
There have been some clear examples of trans-acting factors reported to regulate common fragile site expression. Inhibition of the checkpoint kinase, ATR, resulted in an increased frequency of breaks at common fragile sites (4). ATR is an essential gene required for the cellular response to DNA damage and replicative stress. BRCA1 has also been shown to be required for maintaining fragile site stability (7). BRCA1 acts downstream of ATR, and the G2/M checkpoint function of BRCA1 was shown to be necessary to ensure stability at common fragile sites. In addition, the Fanconi anaemia
(FA) pathway has also been implicated (8). In response to DNA damage, FA proteins associate with BRCA1 and another repair protein, RAD51, and are activated during S-
2 phase. Finally, the structural maintenance of chromosomes (SMC) proteins Smc1 and
Smc3, which are part of the cohesin complex that holds sister chromatids together, are also involved in repair of double strand breaks (DSBs) and have been shown to regulate fragile site stability in an ATR dependent manner (9). A relationship between fragile sites and replication timing has been known for some time. Fragile sites FRA3B, FRA16D and
FRA7H have all been shown to be late replicating regions of the genome (10-12). It has been proposed that the AT-rich sequences at common fragile sites contribute to stalling of replication forks during S-phase. Further delay in replication (in response to aphidicolin in vitro or other stresses in vivo) could then result in non-replicated or
‘fragile’ regions that escape the normal replication checkpoint and they are prone to breakage.
1.2 FRA3B: the most active common fragile site
Several laboratories have cloned and characterized portions of the genomic regions in the vicinity of the most active fragile site, FRA3B (13-15) 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 (16). 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 (13-15). The search culminated in 1996, in the discovery of the very large (1.6 Mb) FHIT gene that encompassed the fragile region,
3 as well as the translocation breakpoint in intron 3 of the FHIT gene (17,18) (Figure
1.1A). Cancer cell lines and primary cancers were shown to harbor deletions within one or both FHIT alleles (17-19) and large fractions of various types of cancers, including upper gastrointestinal tract, colon, cervical, lung and breast cancers, expressed reduced or no Fhit protein (20). 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 (21, 22) by 2006 there were nearly 600 reports concerning alteration of the FHIT gene or protein in cancers arising in almost every organ of the body. Ten 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.
1.3 WWOX tumor suppressor gene structure and expression in cancer
The WWOX gene spans a genomic region of more than half a megabase located at
16q23.3-24.1, a chromosome region commonly involved in LOH in many different types of cancer (Figure 1.1B); it is composed of nine exons, encoding a cDNA of 1245 bp
(23). Studies of oesophageal 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 24). 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 (23-26).
4
3p A 26 25 p14.2p 14 .2 FH I T FHIT 24.3 3' 3' 24.1 23 22 21.3 pSV2neo integrations 21.1 deletions 14.2 hybrid breaks 13 HPV 16 integration 12 t(3;8) translocation
11.1 11.2 5' FRA3FR A 3 B B 5 '
B WWOX 16q q23.3-q24.1 5 '
deletions
FRA16D 3 '
Figure 1.1 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. 5
WWOX mRNA expression is reduced in a series of human breast tumors and breast cancer cell lines (25-27). 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 (28). Reduced Wwox staining was found more frequently in ER (-) or scanty positive tumors. In adjacent normal tissue, reduced Wwox staining was observed in 32.9% of cases. All cases that showed Wwox staining in fewer 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 affected by chemotherapeutic drugs. Nunez et al
(29) 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 (30). 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 adjacent-to-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, LOH in the WWOX locus
6 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 24).
1.4 Epigenetic regulation of WWOX expression in cancer
Expression of many tumor suppressor genes is down-regulated in cancer by epigenetic mechanisms; CpG rich areas (islands) of gene promoter regions are frequently methylated in cancer (for review, 31), 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 (32-35).
We inspected an interval including 1 kb upstream and downstream from the
WWOX transcription initiation site for in silico CpG island identification. 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 -183 to +177 has the highest percentage (~70%) of CpG sites (Figure 1.2).
Ishii et al reported a possible role of epigenetic mechanisms in WWOX transcriptional regulation (36). Specifically treatment of K562 leukemia cells with 5-Aza-
2-deoxycytidine (demethylating agent) and depsipeptide (HDAC inhibitor) increased
WWOX mRNA expression. Kuroki et al (37) studied WWOX promoter methylation status in pancreatic cancer and reported methylation in the region –148 to –37 respective to the transcription initiation site (+1) in pancreatic adenocarcinomas; WWOX promoter
7 methylation status was associated with lack of expression and after treatment with 5-Aza- deoxycytidine in Hs766T cells, Wwox was re-expressed.
promoter exon 1 intron 1 CpGs 16 18 32
S p S p S p ATG 1 1 1
- 406 - 216- 200 - 183 - 59 +1 +126 +177 +232
Highest % of CpGs
Figure 1.2 Wwox regulatory region. The highest percentage of CpG sites is from -183 to +177 relative to the transcription initiation site. 3 CpGs in the promoter region correspond to Sp1 binding sites.
8 1.5 WWOX protein structure and function in cancer
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 (23). WW domains are globular domains consisting of
~40 amino acids, of which two tryptophans and an invariant proline are highly conserved
(38). 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 it has been suggested that WW domain interactions can be modulated by tyrosine phosphorylation
(39). 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 alpha-carbon 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 (40). 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 (41) recently reported that Wwox is upregulated in COS7 fibroblasts and DU145 prostate cancer cells after 17-β-estradiol treatment.
Aqeilan et al (42) 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
9 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 AP2γ via its first WW domain (43). Ap2γ encodes a transcription factor and is frequently up-regulated in breast carcinomas. Wwox overexpression triggered redistribution of nuclear Ap2γ to the cytoplasm, hence suppressing its transactivating function. Interestingly, point mutation in the terminal tyrosine of the PPPY motif abrogates the interactions.
In addition it has been shown that Wwox physically associates with the full-length
ErbB-4 via its first WW domain (44). Coexpression of WWOX and ErbB-4 in HeLa cells followed by treatment with TPA results in the retention of ErbB4 in the cytoplasm.
Moreover, in MCF-7 breast carcinoma cells, expressing high levels of endogenous
Wwox, endogenous ErbB-4 is also retained in the cytoplasm. A mutant form of Wwox lacking interaction with ErbB-4 has no effect on this coactivation of ErbB4. Furthermore,
Wwox is able to inhibit coactivation of p73 by YAP. These data indicate that WWOX antagonizes the function of YAP by competing for interaction with ErbB4 and other targets and thus affects its transcriptional activity.
Wwox-p73, -Ap2γ and -ErbB4 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.
10
CHAPTER 2
WWOX IS EPIGENETICALLY REGULATED IN CANCER
2.1 Abstract
This study aimed to (a) determine if DNA methylation is a mechanism of WWOX and
FHIT inactivation in lung, breast and bladder cancers; (b) examine distinct methylation patterns in neoplastic and adjacent tissues and (c) seek correlation of methylation patterns with disease status. Protein expression was detected by immunohistochemistry, and methylation status by methylation specific PCR and sequencing, in lung squamous cell carcinomas and adjacent tissues, invasive breast carcinomas, adjacent tissues and normal mammary tissues and bladder transitional cell carcinomas. Wwox and Fhit expression was reduced in cancers in association with hypermethylation. Differential patterns of
WWOX and FHIT 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 adjacent and normal tissue. Differential methylation in cancerous vs. adjacent tissues, suggests that WWOX and FHIT hypermethylation analyses could enrich a panel of DNA methylation markers.
11 2.2 Introduction
WWOX (WW domain containing oxidoreductase), is located at a common fragile region, FRA16D, on chromosome 16q23.3 (23) and exhibits genomic alterations in lung, breast, ovary and esophageal cancers (45-47). Point mutations in the gene are infrequent but deletions occur frequently (23,25,48). A recent study showed reduced Wwox expression in 63% of invasive breast tumors and 33% of adjacent normal tissues and
Wwox and Fhit expression was strongly correlated, suggesting that genes at common fragile sites are likely to be coordinately inactivated in cancer (28). The fragile histidine triad (FHIT) gene at chromosome 3p14.2, encodes a tumor suppressor and undergoes frequent hemi- and homozygous deletion in various types of cancer (17). Expression of
Fhit protein is reduced or absent in the majority of human cancers (20). It has been shown that Fhit expression is lost in most lung SCCs, especially those from tobacco smokers
(19), is reduced in invasive breast carcinomas (28,49) and in transitional cell carcinomas of the bladder (50).
The absence of point mutations in WWOX and FHIT genes in cancers with highly reduced protein levels, suggests that other mechanisms, in addition to deletions, might regulate their expression. Methylation of CpG islands, in 5’ regulatory regions of genes, has been associated with transcriptional inactivation of a number of suppressor genes in lung, breast and bladder cancers. Indeed, previous studies have shown that methylation is a mechanism of FHIT inactivation in breast (51), bladder (52) and non-small lung cancer
(53,54). A recent study showed that WWOX promoter hypermethylation was detected in
12 pancreatic cancer cell lines and primary tumors and treatment with the demethylating agent 5-aza-2'-deoxycytidine caused an increase in Wwox expression (37).
We have been interested in mechanisms that control fragile gene expression and inactivation in human cancers and have assessed lung, breast and bladder cancers for control of expression of these genes by methylation of 5’ CpG islands.
2.3 Materials and methods
2.3.1 Patient samples
Frozen lung and breast tumors, adjacent non neoplastic and normal tissue samples, and corresponding fixed archived tissues, were used in this study. Specifically,
24 lung squamous cell carcinomas and corresponding non cancerous adjacent tissues, 28 invasive breast carcinomas, adjacent non neoplastic mammary tissues and 26 normal mammary tissues were obtained from the Kimmel Cancer Center Pathology Core Facility at Thomas Jefferson University, after approval by the Institutional Review Board. DNAs and unstained paraffin sections from 27 transurethrally resected transitional carcinomas
(TCCs) of the bladder and 7 non neoplastic urinary bladder tissues were obtained from the Department of Urology, Thomas Jefferson University. Peripheral blood from a healthy donor served as an internal control.
13 2.3.2 In silico CpG island identification
We used an interval including 1 kb upstream and downstream from the transcription initiation site for in silico CpG island identification (using the CpG plot program http://www.ebi.ac.uk/emboss/ cpgplot) with the following parameters: 100 bp length,
50% C+G and 0.6 observed/expected ratio).
2.3.3 Methylation specific PCR (MSP)
Genomic DNA was extracted from tissues and bisulfite modification of DNA was performed as reported (55). Modified DNA was purified using Wizard DNA purification resin (Promega Corp., Madison, WI). Specific primers, for amplifying methylated or unmethylated products, were designed for the WWOX promoter and exon 1, and for
FHIT intron 1 (Table 2.1). Specific primers covering CpG sites were designed to distinguish between methylated and unmethylated sequences of bisulfite-modified DNA
(supplemental table). The genomic location of MF1 and UF1 primers is -387, MR1 and
UR1 primers -41, MF2 and UF2 primers +72 and MR2 and UR2 primers +162, relative to the start site (+1) of exon 1 of the WWOX gene. For the FHIT gene, the genomic location of MF, UF and MR, UR primers is 39 and 112 relative to the start site of intron 1 of the FHIT gene.
14 Annealing Product Primer Sequence temperature (oC) size (bp)
WWOX MSP 58 347 MF1 5’- TATGGGCGTCGTTTTTTTAGTT-3’ MR1 5’- CAATCTCCGCAATATCGCGACA-3’ Prom oter 58 347 UF1 5’- TATGGGTGTTGTTTTTTTAGTT-3’ UR1 5’- CAATCTCCACAATATCACAACA-3’
MF2 5’- GCGAGTGGATTCGGTAGCGGGCGA-3’ 62 91 Exon 1 MR2 5’- CCGTATCGTCCAACCCCGCGT-3’ UF2 5’- GTGAGTGGATTTGGTAGTGGGTGA-3’ 62 91 UR2 5’- CCATATCATCCAACCCCACAT-3’ Genomic bisulfite sequencing F1 5’- TAGAAGTTTAGGATAATAGTAT-3’ 58 767 First PCR R1 5’- CTCCTTAACAATTACTTTCACT-3’ 58 690 Nest ed F1 5’- TAGAAGTTTAGGATAATAGTAT-3’ PCR R2 5’-TAAACTATACAAAATCCCAAAT-3’ S1 5’- TAGAAGTTTAGGATAATAGTAT-3’ 55 55 Sequencing S2 5’-GTTTTTGTAGGATTGGTTAGAA -3’ S3 5’-TAAACTATACAAAATCCCAAAT -3’ 58 S4 5’-CTACCATAACTAACTATAAA -3’ 55 FHIT M SP MF 5’- TTGGGGCGCGGGTTTGGGTTTTTACGC-3’ 68 74 Methyl ated MR 5’- CGTAAACGACGCCGACCCCACTA- U3’F 5’- TTGGGGTGTGGGTTTGGGTTTTTATG-3’ 68 74 Unme thylated UR 5’- CATAAACAACACCAACCCCACTA-3’
Genomic bisulfite sequencing Region 1 F1 5’- GAAAAAGTTAAAGATTGTGCGA-3’ 55 402 First PCR R1 5’- AAACGACGCCGACCCCACTAAA-3’ Nested F2 5’-AGTTGTGTTTTGTGGTTAGTGTTTTT-3’ 65 350 PCR R2 5’-AAACTTACCTCCCCGCCCCTAC-3’ Sequencing S1 5’-TAGGGTTATTGTTATTATTATGGT -3’ 62
Region 2 F3 5’- GTTATTTAGTGGGTATATTTT-3’ 54 398 First PCR R3 5’- CCCCAAAACCAAAAACTATA-3’ Nested F4 5’- GTTATTTAGTGGGTATATTTT-3’ 54 347
PCR R4 5’- TACCTCAATTTCCCCAATATA-3’ Seque ncing S1 5’-TAGGGTTATTGTTATTATTATGGT -3’ 55
Table 2.1 WWOX and FHIT primers for MSP and genomic bisulfite sequencing
15 2.3.4 Bisulfite treatment and sequencing
Bisulfite modified genomic DNA (2 µl) was amplified for sequencing, using primers as indicated in supplemental table and products were analyzed on 1% agarose gels.
Sequencing of the amplified products was accomplished by BigDye terminator reaction chemistry for analysis on the ABI Prism 377 (Applied Biosystems, Foster City, CA).
2.3.5 Immunohistochemistry
Immunohistochemical studies were performed as described using rabbit polyclonal anti-
Fhit (Zymed, S. San Francisco, CA) and anti-Wwox as primary sera (Guler et al., 2004)
Detection was with strepavidin-biotin complex, using the LSAB2 system (DAKO) with diaminobenzidine as chromogen. In evaluation of Fhit and Wwox staining, intensity and extent of staining were scored. Intensity was graded as: 1, loss of staining, 2, moderate staining and 3, intense staining. Extent of staining was graded as: 1, <10%; 2, 11-25%; 3,
26-50%; 4, 51-75%; 5, 76-100%. Total staining scores were calculated by multiplying intensity and extent scores. Cases with staining scores between 12 and 15 were considered without loss, scores of 6-11, moderately reduced and 1-5 scores, highly reduced expression.
2.4 Results
2.4.1 Wwox and Fhit expression.
We investigated Wwox and Fhit expression levels in the three types of cancer in a subset of our samples in order to verify accordance with previous results (28,48,50). Slides from archived tissues were available for a subset of the neoplastic and non neoplastic tissues,
16 including 6 lung tumors and adjacent tissues, 5 breast cancers and 7 bladder TCCs. The lung tumors, 3 squamous and 3 adenosquamous (ASCCs) cell carcinomas, showed highly reduced Fhit expression (see Figure 2.1A-B, for examples; Table 2.2 for summary);
Wwox expression was moderately reduced in 2 and highly reduced in 4 cases (Figure
2.1C, for example). Wwox and Fhit exhibited similar patterns in adenosquamous tumors; in all 3, the adenocarcinomatous differentiation regions showed stronger staining (Figure
2.1B). In adjacent normal lung tissue, bronchial epithelial cell layers and alveolar macrophages showed strong expression of both Wwox and Fhit (Figure 2.1D). Five invasive ductal carcinomas (IDCs) of the breast were evaluated for Fhit and Wwox expression. One tumor showed highly reduced Fhit expression (Figure 2.1E), 2 had moderately reduced (Figure 2.1F) and 2 cases showed strong expression of Fhit
(summary in Table 2.1). Wwox was moderately reduced in 4 breast cancers (Figure
2.1G, for example) and strongly positive in 1 (Figure 2.1H). Among 7 bladder TCCs, 3 showed highly reduced (Figure 2.1I), 3 moderately reduced (Figure 2.1J) and 1 case strong expression of Wwox (Figure 1K, summarized in Table 2.2). There was adjacent normal transitional epithelium in the same sections in four cases; normal epithelium showed strong cytoplasmic Wwox expression in all 4 (Figure 2.1L, for example). Fhit was highly reduced in 3 bladder cancers, moderately reduced in 2 and highly expressed in
2 (56; summarized in Table 2.2).
17 Figure 2.1 Immunohistochemical detection of Fhit and Wwox expression in lung, breast and bladder tissues. A, Highly reduced Fhit staining in SCC of lung; B, in an adenosquamous carcinoma, only tumor areas with adenoid differentiation showed strong cytoplasmic Fhit reaction; the squamous portion exhibited highly reduced expression; C, reduced expression of Wwox in lung SCC; D, Wwox expression in normal lung tissue; bronchial epithelial cells and alveolar macrophages showed strong cytoplasmic expression; E. highly reduced Fhit staining in breast IDC; residual normal ductal structures showed strong cytoplasmic reaction; F, moderately reduced Fhit staining in an invasive breast tumor with both ductal and lobular features; G, moderately reduced
Wwox reaction in the same tumor; both normal ductal structures showed strong cytoplasmic reaction in the luminal epithelial layer; H, strong Wwox expression in an
IDC, adjacent normal breast lobules showed strong cytoplasmic expression; I, highly reduced; J, moderately reduced and K, strong cytoplasmic Wwox expression in bladder
TCC; L, strong cytoplasmic expression of Wwox in adjacent normal epithelium of bladder
18
19
WWOX FHIT Tiss ue Samples Diagnosis expression* methylation ** expression* methylation ** 296T SCC -- 24% -- 24%
349T SCC + 35% -- 31% 370T SCC -- 24% -- 21% lung 491T SCC -- 19% -- 28% 523T SCC + 19% -- 28% 564T SCC -- 21% -- 24%
81T IDC + 27% -- 100% 491T IDC + 36% + 22% brea st 539T IDC + 13% + 32% 108T IDC + 20% ++ 7% 194T IDC ++ 6% ++ 3%
BLC22 TCC + 24% -- 100% BLC25 TCC -- 24% -- 21% BLC4 TCC -- 56% -- 21% bladder BLC21 TCC -- 18% + 4% BLC26 TCC + 11% + 10%
BLC20 TCC + 11% ++ 0% BLC15 TCC ++ 0% ++ 0%
* protein expression assesed by immunohistochemistry: --, staining score of 1-5 represents highly reduced expression; + score of 6-11 represents reduced expression; ++ score of 12-15 represents a normal level of expression
** % methylated CpG sites/ total CpG sites
Table 2.2 Summary of FHIT and WWOX methylation status and expression
20 2.4.2 FHIT and WWOX CpG islands.
Epigenetic alterations of chromatin, such as DNA methylation and histone modification, are associated with transcriptional inactivation of many tumor suppressor genes (57-59).
The methylation pattern of the FHIT 5’ CpG island was described by Tanaka et al (57) in a study showing that the region between nucleotides 195 and 283 (intron 1) is sensitive to methylation. Ensuing studies of FHIT methylation in non small lung, breast and bladder cancers (52,53) focused on assessment of methylation of intron 1 CpGs by MSP.
There have not been detailed studies of WWOX transcriptional regulation by epigenetic mechanisms, including CpG island methylation. Bednarek et al (23) suggested the presence of a CpG island in exon 1 of the WWOX gene. A region 406 bp upstream of the WWOX trascription start site, exon 1 and a part of intron 1 (the first 125 bp) exhibited
66% C+G residues; specifically, the region from -183 to +177 has the highest percentage
(~70%) of CpG sites.
2.4.3 WWOX and FHIT methylation status.
The MSP method was used to examine CpG islands in DNAs from the lung SCCs and adjacent non-malignant lung tissues, breast IDCs, adjacent non neoplastic mammary tissues and normal mammary tissues, TCCs of the bladder and non neoplastic urothelium tissues. Representative results of MSP amplification of the WWOX promoter (347 bp),
WWOX exon 1 (91 bp) and FHIT intron 1 (74 bp) are displayed in Figure 2.2A, B, C.
The CpG islands in the WWOX promoter region were methylated in 62.5 % of the lung
SCCs, in WWOX exon 1 in 46% of the cases and FHIT intron 1 in 38% of the cases
(Figure 2.2D). Only WWOX exon 1 was highly methylated in the adjacent non-
21 neoplastic lung tissues. In breast IDCs, the WWOX promoter and exon 1 were methylated in 53% and 42% of cases respectively, and FHIT intron 1 was methylated more frequently (57%). A large fraction of non neoplastic mammary tissues adjacent to cancer
(46%) showed WWOX promoter methylation. WWOX and FHIT DNA of normal mammary tissues were completely unmethylated in all cases. In TCCs of the bladder, the
WWOX promoter and exon 1 were infrequently methylated (29% and 14% of cases, respectively), while FHIT intron 1 was methylated in 11% of cases. The non neoplastic bladder tissue DNA was unmethylated for both genes (Figure 2.2D). In summary,
WWOX and FHIT regulatory region CpGs are highly methylated in lung and breast cancer DNAs, less so in bladder cancers.
2.4.4 Correlation of WWOX and FHIT methylation status and association with clinicopathological features.
We sought correlation of the methylation status of the fragile genes with clinicopathological features (sex, age, clinical stage, lymph node metastasis and histology). Fisher’s exact test was used to examine the association between WWOX and
FHIT methylation and clinicopathological features. A two-sided P-value of less than 0.05 was considered statistically significant.
22 A 296T 296N BCL4 BCL17 137T 142N M U M U M U M U M U M U 347 bp
B 296T 296N BCL4 BCL17 137T 142N M U M U M U M U M U M U 91 bp
C 296T 296N BCL4 BCL22 137T 142N M U M U M U M U M U M U 74 bp
D
Tissues Description WWOX promoter WWOX exon 1 FHIT intron 1 Squamous cell carcinomas 15/24 (62.5%) 13/24 (46%) 9/24 (38%) Lung Non neoplastic adjacent tissues 1/24 (4%) 7/24 (29%) 2/24 (8.3%) Invasive breast carcinomas 15/28 (53%) 12/28 (42%) 16/28 (57%) Non neoplastic adjacent tissues 13/28 (46%) 1/28 (3%) 4/28 (14%) Breast Normal tissues 0/26 (0%) 0/26 (0%) 0/26 (0%) Transitional cell carcinomas 8/27 (29%) 4/27 (14%) 3/27 (11%) Bladder Non neoplastic adjacent tissues 0/7 (0%) 0/7 (0%) 0/7 (0%)
Figure 2.2 MSP amplification in WWOX and FHIT regulatory regions. MSP analysis of 296T (lung SCC), 296N (normal lung), BLC4, BLC17, BLC22 (TCC), 137T (IDC) and 142N (non neoplastic mammary tissue) in: A, The promoter region of the WWOX gene (347 bp); B, exon1 of the WWOX gene (91 bp); and C, intron 1 of the FHIT gene (74 bp). M and U lanes represent amplified products of primers recognizing methylated and unmethylated sequences, respectively. D, High frequency WWOX and FHIT methylation in lung and breast but lower frequency in bladder tissues. High frequency of methylation was detected in the WWOX promoter, exon 1 and FHIT intron 1 in SCCs but the corresponding adjacent normal tissues showed a high frequency of methylation only in WWOX exon 1.
23
In lung SCCs, WWOX promoter and exon 1 methylation was observed very frequently in male patients (13/14) (p= 0.011) and most of the cancers without methylation of WWOX promoter and/or exon 1 were from females (8/10). WWOX and
FHIT methylation did not correlate significantly with known clinicopathological characteristics of breast cancers. On the other hand 11 of 16 breast cancers methylated in
WWOX exon 1 were also methylated in FHIT intron 1, showing a strong positive association between WWOX exon 1 and FHIT intron 1 (p=0.002) methylation, stronger than the association of WWOX promoter and FHIT intron 1 methylation (p=0.02).
2.4.5 Methylation patterns assessed by sequencing.
To obtain detailed methylation profiles, we analysed 62 CpG sites spanning the WWOX promoter, exon 1 and a small region of intron 1, and 29 CpG sites in the FHIT promoter, exon 1 and intron 1, by sequencing amplified products of bisulfite treated DNAs
(representative results summarized in Figure 2.3).
19-35% of the WWOX CpG sites in lung SCCs were methylated: two small groups of hypermethylated CpG sites in the promoter and one group in exon 1 were consistently methylated. In FHIT intron 1 two groups of CpGs were consistently methylated with 21-31% of CpGs methylated. The adjacent non cancerous lung tissue
DNAs exhibited methylation in WWOX exon 1, but no FHIT methylation.
In breast carcinomas 5-36% of the WWOX CpG sites were methylated, creating two CpG islands in the promoter and one in exon 1; the number of methylated FHIT CpG sites ranged from 3-100%, but intron 1 CpGs were always methylated. We also observed
24 methylation in some CpG sites of WWOX promoter and FHIT intron 1 in adjacent non neoplastic mammary tissues. WWOX and FHIT regulatory regions were unmethylated in normal mammary tissues.
In some TCCs, WWOX was methylated at 15-56% of CpG sites, mainly in the promoter and exon 1. We did not detect a specific methylation pattern for the FHIT gene in bladder cancer and WWOX and FHIT DNA methylation may be an infrequent mechanism of inactivation in bladder cancer.
2.4.6 WWOX and FHIT promoter methylation in cancers correlates with reduced protein expression.
To examine the relationship between WWOX and FHIT methylation and protein expression, we assessed levels of expression of the proteins by immunohistochemistry in lung SCCs, breast IDCs and TCCs (Table 2.2). In lung SCCs there was a correlation between methylation of WWOX and FHIT promoters and protein expression, confirming that promoter methylation is an important inactivation mechanism for Wwox and Fhit in lung cancer. WWOX and FHIT regulatory regions were methylated in all the lung tumors and protein expression was highly reduced. WWOX and FHIT DNA methylation was also associated with reduced protein expression in breast cancer, with most invasive breast carcinomas showing reduced expression of both. WWOX and FHIT DNA methylation was also associated with reduced protein expression in bladder cancer, with most TCCs showing reduced expression for both (Table 2.2). WWOX and FHIT CpGs were unmethylated in normal tissues and both proteins were expressed.
25 Figure 2.3 Methylation patterns in WWOX and FHIT 5’ regulatory regions.
Sequencing of bisulfite treated DNA revealed different methylation patterns for WWOX
(A) and FHIT (B) genes. Black squares represent methylated CpG sites, white squares are unmethylated CpGs sites and grey are hemimethylated CpGs. The first WWOX CpG site is located -343 bp relative to the start of exon 1 and the first FHIT CpG site is located -94 bp relative to the start of exon 1.
26
27
2.5 Discussion
Because genes at common fragile sites are likely to be inactivated by similar mechanisms (17, 60), we have been interested in coordinate inactivation of fragile genes in cancer and have shown that loss or reduction of Fhit and Wwox expression is tightly associated in invasive breast cancers (28). Thus, we investigated the mechanism of inactivation of the two genes in three types of cancer, lung, breast and bladder. If the two genes should be coordinately inactivated in a large fraction of several types of cancers, they might serve as biomarkers for multiple cancer types, and eventually as surrogate markers for prevention and therapeutic trials.
We confirmed that WWOX and FHIT are frequently inactivated in lung, breast and bladder cancer and that expression is reduced or lost coordinately with promoter region methylation. Wwox expression was highly reduced in lung cancer and reduced in breast and bladder cancers. Yendamuri et al (2003) showed that WWOX alterations occur in a significant fraction of lung cancers; in addition, in a previous study of 94 invasive breast cancers we observed reduced Wwox and Fhit expression in 63% and 55% of the cases (Guler et al., 2004). Thus, WWOX and FHIT, two genes that are located in fragile sites in different chromosomes, are coordinately inactivated in lung and breast cancer. An interesting aspect of these results is that WWOX and FHIT methylation patterns for the three types of cancers were relatively consistent for a given type of cancer, but differed among cancers from different organs. The pattern that was shared, by tumors with hypermethylation, was consistent methylation of two regions in the promoter
28 and one region in exon 1 of WWOX and a region in FHIT intron 1, in accord with previous methylation studies for the FHIT gene (Esteller et al., 1999; Yang et al., 2002;
McGregor et al., 2002).
Sparse methylation in WWOX exon 1 in non neoplastic lung tissue adjacent to cancer may be an early marker of carcinogen exposure; perhaps when extent of methylation reaches a critical level (~20%), or when an additional epigenetic alteration, such as Fhit inactivation occurs, preneoplastic changes appear.
Interestingly, WWOX promoter and FHIT intron 1 were methylated in DNA from non neoplastic mammary tissues adjacent to cancer, but not in mammary tissues from non cancer patients. In this study we showed that FHIT regulatory region CpGs outside intron
1 could also be methylated, as shown for the promoter and exon 1 in some breast cancers.
WWOX exon 1 methylation differentiates breast cancer from non neoplastic tissue adjacent to cancer and normal mammary tissue.
FHIT and WWOX methylation was less frequent in bladder cancer, and when it occurred, the number of methylated FHIT CpGs was high. According to WWOX methylation status it seemed that there are two different bladder tumor groups. In one group WWOX promoter region is heavily methylated and in the other WWOX is not methylated. The findings in breast and bladder cancer showed that comprehensive examination of entire 5’ regions of a gene may be useful in identifying regulatory regions correlating with epigenetically mediated loss of gene function, and in molecular marker studies to distinguish neoplastic from normal cells. 29 methylated CpG sites, as observed in FHIT for several breast and bladder cancers, is among the highest number of methylated CpGs reported for a regulatory region of a tumor suppressor gene. As we 29 found that methylation in FHIT intron 1 was sufficient for total loss of Fhit expression in lung and some bladder tumors, it would be interesting to investigate whether heavy methylation of the FHIT gene is related to morphological features, environmental exposure and clinical outcome, or could affect response to methylation inhibitors.
The application of molecular markers specific for lung, breast or bladder cancer offers possibilities for early detection (61-64). WWOX can be included as a biomarker for early detection of lung cancer, as well as for monitoring the response to chemopreventive agents. WWOX exon 1 MSP may be used as a marker for early lung carcinogenesis while WWOX promoter MSP differentiates lung cancer from adjacent non neoplastic lung tissue. The high frequency of WWOX and FHIT methylation in association with loss of expression, suggests that these fragile genes could enrich a hypermethylation panel for lung, breast, bladder and probably other cancers.\
Publication: Iliopoulos D, Guler G, Han SY, Johnston D, Druck T, McCorkell KA, Palazzo J, McCue PA, Baffa R, Huebner K. Fragile genes as biomarkers: epigenetic control of WWOX and FHIT in lung, breast and bladder cancer. Oncogene. 24: 1625-1633 (2005).
30
CHAPTER 3
ADENOVIRAL WWOX RESTORATION SUPPRESSES TUMORIGENICITY IN
VITRO AND IN VIVO
3.1 Abstract
The WWOX (WW domain containing oxidoreductase) gene at common fragile site,
FRA16D, is altered in many types of cancer, including lung cancer. We have examined the tumor suppressor function of WWOX in pre-clinical lung cancer models. The
WWOX gene was expressed in lung cancer cell lines through recombinant adenovirus infection (Ad-WWOX), and through a drug [ponasterone A, (ponA)]-inducible system.
After WWOX restoration in vitro, endogenous Wwox protein-negative cell lines (A549,
H460, and H1299) underwent apoptosis through activation of the intrinsic apoptotic caspase cascade in A549 and H460 cells. Ectopic expression of Wwox caused dramatic suppression of tumorigenicity of A549, H460 and H1299 cells in nude mice, after Ad-
WWOX infection, and after ponA induction of Wwox expression in H1299 lung cancer cells. Tumorigenicity and in vitro growth of U2020 lung cancer cells was unaffected by
Wwox overexpression. This study confirms that WWOX is a tumor suppressor gene, and is highly effective in gene therapy of lung cancer xenografts, whether transduced by adenovirus or re-expressed through drug therapy.
31
3.2 Introduction
Lung cancer is the first cause of cancer mortality in the United States (65), with an incidence of ~170,000 new cases/year in the United States (65), and mortality is very high. Non Small Cell Lung Cancer (NSCLC) comprises ~80% of lung cancers. Surgery remains the main therapy for NSCLC, but a large fraction of patients cannot undergo curative resection. Despite new drugs and therapeutic regimens, the prognosis for lung cancer patients has not significantly changed in the last 10 years. Recombinant virus gene therapy has been investigated in lung cancer patients; Adenovirus and retrovirus encoding wild-type p53 have been injected intra-tumorally in lung cancer clinical trials
(66-70). Recombinant adenovirus injection in lung cancer phase I studies (71) has demonstrated safety and feasibility and phase I/II clinical trials are currently recruiting patients to evaluate toxicity and efficacy of gene therapy with recombinant adenoviruses
(72).
Lung cancer is associated with early loss of expression of the FHIT gene (Fragile
Histidine Triad Gene) (73) at fragile site FRA3B (17). Fragile regions are particularly susceptible to damage on exposure to environmental carcinogens, which are etiological factors in lung cancer. Recently, Yendamuri et al. (48) have demonstrated that the
WWOX gene is also altered in a fraction of NSCLCs. WWOX is located at fragile site
FRA16D (23) and encodes a 414-amino acid protein with two WW domains and a short chain dehydrogenase domain. WW domains are protein-protein interaction domains and
Wwox interactors with important signaling roles in normal epithelial cells have been
32 identified. Wwox interacts with p73 and can trigger redistribution of nuclear p73 to the cytoplasm, suppressing its transcriptional activity (42). Wwox also interacts with Ap2 transcription factors with roles in cell proliferation (43). Most recently, Wwox has been reported to compete with Yap protein for binding to the intracellular ErbB4 domain, a transcriptional activator (44). Thus, the Wwox pathway includes a number of downstream signaling proteins that may also serve as cancer therapeutic targets.
The WWOX gene is altered in many types of cancer, including breast, ovary, prostate, bladder, esophagus, and pancreas (26,37,46-47,74). In NSCLC, transcripts missing WWOX exons were detected in 26% of tumors and in 5 of 8 cell lines (48).
WWOX allele loss occurred in 37% of tumors and the promoter is hypermethylated in
62.5% of squamous cell lung carcinomas (48, 74). To investigate tumor suppression in lung cancer, we studied in vitro and in vivo effects of Wwox protein expression, in
Wwox-negative (A549, H460, H1299) and positive lung cancer cells (U2020), by infection with adenovirus carrying the WWOX gene; H1299 cells were also stably transfected with an inducible Wwox expression vector, which allows induction of near physiologic levels of protein. Wwox restoration effectively induced apoptosis in vitro and suppressed lung cancer tumorigenicity in nude mice, with no effect on lung cancer cells constitutively expressing the Wwox protein.
3.3 Materials and methods
3.3.1 Cell culture.
Wwox-negative A549, H460, H1299 and positive U2020 lung cancer cell lines from
American Type Culture Collection (Manassas, VA) were maintained in RPMI 1640 with 33 10% fetal bovine serum. HEK-293 Cym-R cells from Qbiogene (Carlsbad, CA) were cultured in DMEM with 10% fetal bovine serum.
3.3.2 Recombinant Adenoviruses and In vitro Transduction.
WWOX cDNA from normal human liver RNA (Ambion, Austin, TX) was reverse transcribed by SuperScript First-Strand Synthesis (Invitrogen, Carlsbad, CA) and double- stranded cDNA prepared by PCR amplification using the following conditions: 95˚ for 3 min, 30 cycles at 94˚C for 30 sec, 65˚C for 60 sec, 72˚C for 30 sec, and 72˚C for 7 min, and WWOX forward 5’- GCCAGGTGCCTCCACAGTCAGCC- 3’ and WWOX reverse
5’- TGTGTGTGCCCATCCGCTCTGAGCTCCAC- 3’ primers. The cDNA was cloned into Adenovator-CMV5(CuO)-IRES-GFP transfer vector (Qbiogene) (12). This vector allows transgene expression driven by the cumate-inducible CMV5(CuO) promoter. An
IRES sequence ensures co-expression of GFP. The recombinant plasmid, Ad-WWOX, was transfected into modified human fetal kidney HEK-293 CymR cells (Qbiogene), constitutively expressing the CymR protein, which represses the CMV5(CuO) promoter and expression of Wwox during packaging and expansion of the WWOX adenovirus.
After 14-21 days, homologous recombination occurred in cells, leading to plaque formation. Plaques were isolated, viruses amplified in HEK-293 CymR cells and purified by CsCl gradient centrifugation. Titers were determined by plaque assay and transgene expression assessed by immunoblot using Wwox monoclonal antibody (21). Cells were transduced with recombinant adenoviruses at increasing multiplicities of infection (MOI) and transduction efficiency determined by visualization of GFP-expressing cells.
34
3.3.3 Inducible WWOX transfectants.
The human WWOX cDNA was cloned into BamHI and EcoRI sites of the pIND vector.
H1299 cells were transfected with 10 µg of pVgRXR vector, which contains the ecdysone nuclear receptor subunits, clones selected and tested for ponA-inducible expression by transient transfection with a reporter plasmid. Clones showing the highest expression were transfected with 10 µg of the pIND-WWOX vector and cultured in zeocin (150 µg/ml) and G418 (1200 µg/ml). H1299/I clones were selected and tested for inducible WWOX expression after ponA (5-10 µM) treatment.
3.3.4 Western blot Analysis.
Protein extraction and immunoblot analysis were performed as described previously (42).
The following primary antisera were used: mouse monoclonal anti-Wwox 1:500, rabbit polyclonal anti-caspase 3, 1:1000 (Cell Signaling, Beverly, MA), rabbit polyclonal anti- caspase 9, 1:200 (Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal anti- caspase 8 (Cell Signaling) 1:1000, rabbit polyclonal anti-PARP 1:1000 (Cell Signaling), rabbit polyclonal anti β-actin 1:1000 (Cell Signaling).
3.3.5 Cell Growth and cell cycle kinetics.
Cells (2 x 105) were infected at MOI 10, 25, 50, 75 and 100 and at 24 h intervals harvested, stained with trypan blue and counted (ViCell counter, Beckman Coulter, 35 Fullerton, CA). For flow cytometry, cells were harvested 5 days after infection, fixed in cold methanol, RNAse treated and stained with propidium iodide (50 µg/ml). Cells were analyzed for DNA content by EPICS-XL scan (Beckman Coulter) using doublet discrimination gating. All analyses were in duplicate.
3.3.6 In vivo studies.
Animal studies were performed according to institutional guidelines. H460, A549, and
U2020 cells were infected in vitro with 100 MOI Ad-WWOX, Ad-GFP or mock infected.
At 24 h after infection, 5x106 viable cells were injected subcutaneously (s.c.) into left flanks of 6-week-old female nude mice (Charles River, MA), 5 mice per infected or control cell line. H1299 cells were infected in vitro with Ad-GFP or Ad-WWOX at MOI
100. H1299/I cells treated with 10 µM ponA (H1299/I+ cells) to induce Wwox expression. Tumorigenic controls were uninduced cells (H1299/I-). Induced (H1299/I, 24 h after ponA) and uninduced cells (107) were injected into 5 nude mice; 5 mice were also injected with Ad-WWOX, Ad-GFP and mock-infected H1299 cells. Tumor diameters were measured every 5 days and tumors weighed after necropsy. Tumor volumes were calculated using the equation V (mm3)= axb2/2; a is the largest diameter and b is the perpendicular diameter.
36 3.3.7 Ex vivo studies.
Protein lysates from tumors of H1299, H1299/I- and H1299/I+ injected mice were evaluated for Wwox expression by immunoblot analysis. Fragments from H1299/I+ tumors were cultured and treated with 10 µM ponA for 2 days to detect expression of inducible Wwox by immunoblot.
3.3.8 Statistical analysis.
Results of in vitro and in vivo experiments were expressed as mean ± SD. Student’s two- sided t test was used to compare values of test and control samples. P<0.05 indicated significant difference.
3.4 RESULTS
3.4.1 Wwox expression in parental and Ad-WWOX infected lung cancer cells.
Immunoblot analysis of proteins of lung cancer cell lines showed that A549,
H460 and H1299 cells did not express endogenous Wwox, whereas Wwox was detected in U2020 cells. Breast cancer MCF-7 cells express abundant endogenous Wwox (19) and served as positive control (Figure 3.1A).
Lung cancer cells were infected with Ad-WWOX or Ad-GFP at MOI 100; the adenoviral transgene was expressed in nearly 100% of cells of each cell line, as assessed by confocal microscopy of GFP fluorescence (data not shown). Immunoblot analysis 72h after infection showed Wwox overexpression in all Ad-WWOX-transduced cells (Figure
3. 1B).
37
A
1 2 3 4 5 WWOX
β- Actin
B
1 2 3 4 5 6 7 8 9 WWOX
β- Actin
Figure 3.1 Expression of Wwox protein. A) Expression of endogenous Wwox is detected in U2020 and MCF7 but not in H1299, H460, or A549 cells (50 µg proteins loaded): lane 1, H1299; lane 2, H460; lane 3, A549; lane 4, U2020; lane 5, MCF-7. B)
Expression of Wwox after infection with Ad-WWOX (25 µg loaded): lane 1, H1299 Ad-
WWOX infected; lane 2, H1299 Ad-GFP infected; lane 3, H1299; lane 4, H460 Ad-
WWOX infected; lane 5, H460Ad-GFP infected; lane 6: H460; lane 7, A549 Ad-WWOX infected; lane 8, A549 Ad-GFP infected; lane 9, A549.
38 3.4.2 Cell cycle kinetics of infected cells.
Cell cycle alterations induced by Wwox overexpression were assessed after infection at several MOIs, with Ad-WWOX or Ad-GFP. A sub-G1 population was observed after Ad-
WWOX infection in A549, H460, and H1299 cells, that do not express endogenous
Wwox, but not in endogenous Wwox-positive U2020 cells. Ad-GFP infection did not modify cell cycle profiles. At 96 h after Ad-WWOX infection (MOI 100), 58% of A549,
94% of H460, and 17% of H1299 cells were in the sub-G1 fraction; 7% of U2020 cells were in sub-G1 (Figure 3.2). Wwox induction of cell death was MOI- and time- dependent (data not shown).
3.4.3 Apoptotic pathways in Wwox-reexpressing cells.
A549, H460, H1299, and U2020 lung cancer cell lines were infected with increasing
MOIs, and the fraction of transduced cells was monitored by confocal microscopy and cell cycle kinetics anayses. Significant differences were observed in cell growth for Ad-
WWOX and Ad-GFP infection, at a range of MOIs, in lung cancer cell lines, A549, H460 and H1299, lacking endogenous Wwox (Figure 3.3A). U2020 cells were unaffected by exogenous Wwox expression.
To study Wwox-induced apoptotic pathways, expression of downstream apoptotic effectors was assessed in vitro. At 96 h after infection, pro-caspase 3 and full-length poly
(ADP-ribose) polymerase-1 (PARP-1) levels were reduced in Ad-WWOX-infected A549 and H460 cells compared to Ad-GFP control cells. In H1299 cells a decrease of full- length PARP-1 was observed. Cleavage of precursors was not observed in infected
U2020 cells (Figure 3.3B).
39 Sub G1: 2.33 % Sub G1: 2.33 % Sub G1: 57.76 %
Sub G1: 3.83% Sub G1: 2.75 % Sub G1: 94.3 %
Sub G1: 17.46 % Sub G1: 1.4 % Sub G1: 2.0 %
Sub G1: 5.7 % Sub G1: 1.37 % Sub G1: 7.39 %
Figure 3.2 Flow cytometry analysis of untreated, Ad-GFP and Ad-WWOX infected cells. Wwox negative A549, H460, H1299 cells undergo apoptosis 5 days after restoration of Wwox expression by Ad-WWOX infection, but U2020 cells are unaffected. Ad-GFP infection did not induce apoptosis.
40
Figure 3.3 Effect of Wwox expression on cell growth in vitro. A) Growth of uninfected, Wwox negative A549, H460, and H1299 cells and after infection with Ad- GFP and Ad-WWOX. B) Immunoblot detection of full-length PARP and pro-caspase 3: lane 1, A549; lane 2, A549/Ad-GFP; lane 3, A549/Ad-WWOX; lane 4, H460; lane 5, H460/Ad-GFP; lane 6, H460/Ad-Wwox; lane 7, H1299; lane 8, H1299/Ad-GFP; lane 9, H1299/Ad-WWOX; lane 10, U2020; lane 11, U2020/Ad-GFP; lane 12, U2020/Ad- WWOX. PARP is cleaved in Wwox negative cell lines when Wwox is restored through Ad-Wwox infection (lanes 3,6,9). Procaspase 3 expression is highly reduced in A549 and H460 (lanes 3,6) but not in H1299 cells after Ad-WWOX infection. In U2020 cells, neither PARP, nor pro-caspase 3 are cleaved after Ad-WWOX infection (lane 12).
41 3.4.4 Effects of conditional Wwox expression in H1299 cells.
H1299/I clone 7 expressed the WWOX transgene only on induction with ponA (Figure
3.4A), and was used in subsequent experiments. Wwox expression increased in a dose- dependent manner after ponA treatment (Figure 3.4B) from 24 to 72 h (Figure 3.4C).
Clone 7 H1299/I- (uninduced) cells were plated and 24 h later (day 1), Wwox expression was induced by 10 µΜ ponA. Maximum expression was observed at day 4 and significantly affected cell proliferation by day 5 (Figure 3.4D), causing reduction in cell numbers and suggesting that Wwox inhibits growth of H1299 cells.
3.4.5 Tumorigenicity of Ad-WWOX-infected lung cancer cell lines.
Nude mice were inoculated with 5 x 106 A549, H460, and U2020 cells infected in vitro at
MOI 100 with Ad-GFP or Ad-WWOX and cultured for 24 h. Uninfected cells served as tumorigenic controls. Tumor growth was completely suppressed in mice inoculated with
Ad-WWOX-infected H460 cells, at 28 days after injection (Figure 3.5A). The average tumor wt for controls (Ad-GFP and untreated H460 cells) at day 28 were 0.61 ± 0.15 g, and 0.64 ± 0.11 g, respectively. At 28 days, 2/5 mice inoculated with Ad-WWOX- infected A549 cells showed no tumors and average tumor wt was 0.08 ±0.03 g, significantly lower (p<0.001) than tumors of Ad-GFP infected A549 (0.81 ± 0.16 g) and mock infected A549 cells (0.86 ± 0.15 g) (Table 3.1). In mice injected with infected
U2020 cells, no tumor growth suppression was observed (Figure 3.5A).
42
Figure 3.4 Inducible expression of Wwox in H1299/I cells. A) Cells were cultured in the presence (+) or absence (-) of 10 µM ponA for 48 h and tested for Wwox expression. Clones 7 and 2, which expressed the transgene only upon induction with ponA, were used in subsequent experiments. Gapdh expression served as loading control. B) H1299/I clone 7 cells incubated in the absence or presence of increasing concentrations of ponA for 48 h: Wwox levels increased in a dose-dependent manner and were quantified by densitometry, normalized to Gapdh expression levels. C) Time course of Wwox induction in H1299/I clone 7 cells after treatment with 10 µM ponA: Wwox levels were quantified by densitometry. D) Effect of 10 µΜ ponA on growth of H1299/I clone 7 cells. On Day 1, ponA was added and maximun Wwox expression was found on day 4. From day 5 the induced cells (H1299/I+) grow significantly more slowly than uninduced cells (H1299/I-) (p<0.001). The experiment was done in triplicate. 43
Figure 3.5 Effect of Wwox expression on tumorigenicity of lung cancer cells. A) Tumor volume of untreated, Ad-GFP- and Ad-WWOX-infected A549, H460 and U2020 lung cancer cells. Restoration of Wwox expression in A549 and H460 cells suppressed tumor growth significantly (p<0.001) compared to Ad-GFP infected cells. B) Tumor volume of untreated, Ad-GFP and Ad-WWOX infected H1299 cells and H1299/I- and H1299/I+ cells. Tumors were suppressed in Ad-WWOX infected H1299 cells and in H1299/I+ cells. C) Examples of tumor formation by uninfected, Ad-GFP and Ad- WWOX infected A549, H1299/I- and H1299/I+ cells. 44 3.4.6 Effect of drug-induced Wwox expression on tumorigenicity.
We next compared tumorigenicity of H1299 cells infected with Ad-WWOX or induced to express Wwox by ponA treatment. Nude mice were inoculated with 1 x 107 cells 24 h after infection with Ad-WWOX or Ad-GFP. Five mice were also injected with 1 x 107 uninduced H1299/I (H1299/I-) and 107 H1299/I+ cells 24 h after ponA treatment. At 28 days after injection, 3/5 and 4/5 mice inoculated with Ad-WWOX-infected H1299 cells and H1299/I+ cells, respectively, displayed no tumors (Figure 3.5B). Average wt of tumors from Ad-WWOX-infected cells (0.10 ± 0.26 g) and H1299/I+ cells (0.21 ± 0.31 g) was significantly reduced compared to tumors from Ad-GFP (1.66 ± 0.28 g), H1299/I-
(1.98 ± 0.41 g) and parental H1299 cells (1.87 ± 1.33 g) (Table 3.1). Thus, Wwox expression, delivered by gene therapy (Ad-WWOX) or by induction of expression of an inactive “endogenous” WWOX gene (H1299/I+), was effective in suppressing lung cancer cell growth in nude mice.
3.4.7 Wwox expression in H1299/I+ explanted tumors. To assess Wwox expression ex vivo, we performed immunoblot analysis of proteins extracted from fragments originating from parental H1299, H1299/I- and H1299/I+ tumors; Wwox expression was not found in any of the tumors (Figure 3.6A). Explanted, cultured fragments from H1299/I+ tumor were examined for retention of inducible WWOX plasmid by treating with ponA and testing for Wwox expression by immunoblot analysis. The detection of Wwox induction in H1299/I+ explants revealed that the WWOX plasmid was present and inducible
(Figure 3.6B), suggesting that the small tumors were derived from inoculated cells that had lost expression of Wwox due to absence of inducer in vivo.
45 WWOX - WWOX +
A549 H460 H1299* U2020
Untreated 0.86±0.15 0.64±0.11 1.87±0.33 0.57±0.09 Ad-GFP 0.81±0.16 0.61±0.15 1.66±0.28 0.55±0.05
Ad-WWOX 0.08±0.03 0.03±0.04 0.10±0.26 0.59±0.03 *H1299/I-: 1.98 ± 0.41 H1299/I+: 0.21 ± 0.31
Table 3.1 Tumor weight (g) ± SD in nude mice
A
1 2 3 4 5 6
Wwox
Gapd h
B 0 5 7.5 10 Pon A (μWΜwo)x
Gapd h
Figure 3.6 Ex vivo analysis of H1299/I- and H1299/I+ cells. A) Protein lysates from H1299 (lane 1), uninduced H1299/I- (lanes 2,3,4) and induced H1299/I+ (lane 5) tumors tested for Wwox expression by immunoblot analysis: Wwox was not expressed in the H1299/I- or H1299/I+ tumors; B) A portion of the H1299I/+ tumor was plated, cultured and cells were treated with ponA. Wwox was re-expressed after 48 h treatment with 10 µM ponA, indicating presence of the inducible WWOX plasmid.
46 3.5 DISCUSSION
New therapeutic strategies are urgently needed for lung cancer treatment. Because genes at common fragile sites are frequently inactivated early in the neoplastic process, especially on exposure to environmental carcinogens, we have been interested in the effect of loss of fragile gene expression in development of cancer and therapeutic effects of their restoration (75). A number of studies have suggested that the fragile Wwox gene is inactivated in a significant fraction of lung cancers (37,48), particularly by promoter hypermethylation (74). Hypermethylation is reversible, a strategy with promise for cancer therapy. Thus, we have determined if restoration of Wwox expression in lung cancer cells lacking expression of endogenous Wwox, would reverse malignancy in spite of numerous cancer-associated genetic alterations that have accumulated in lung cancer cell lines. We have restored Wwox expression in 4 lung cancer cell lines by infection with Ad-WWOX and observed dramatic loss of tumorigenicity of the lung cancer cells that lacked endogenous Wwox.
Introduction of the WWOX gene in the 3 Wwox-negative cell lines resulted in induction of apoptosis in vitro, as shown by the fraction of cells with sub-G1 DNA content and by suppression of cell growth in culture. The fraction of Ad-WWOX infected
H1299 cells with sub-G1 DNA content, was lower than for the other two WWOX- negative cell lines, possibly because apoptosis may occur later after restoration of Wwox expression in H1299 cells. The U2020 lung cancer cells expressing endogenous Wwox were not affected by overexpression of Wwox, suggesting that normal Wwox expressing lung cells would be unaffected by Wwox overexpression after WWOX gene therapy.
47 Growth of all 3 lung cancer cells in vitro was adversely affected by overexpression of
Wwox after virus infection or ponA induction, as shown by the downturn in cell number after a few days of Wwox overexpression. It will be interesting to examine Wwox binding to known interacting proteins at days 2 through 5 in these in vitro overexpression cultures to define the signal events directly downstream of Wwox expression, after
WWOX infection or induction.
We observed efficient suppression of in vivo tumorigenicity of lung cancer cell lines by Ad-WWOX transduction in 3 WWOX-negative lung cancer cell lines, and by induction of Wwox expression in stably transfected H12990 lung cancer cells. The tumorigenicity of the aggressive H460 cell line was completely suppressed by Ad-
WWOX treatment at 28 days after injection. A significant reduction in tumor occurrence and size was observed in animals injected with WWOX-transduced A549 and H1299 cells. The results suggest that Wwox loss may play an important role in the pathogenesis of lung cancer. It is interesting that both methods of Wwox restoration in H1299 cells appeared to result in more dramatic effects in vivo than in vitro, possibly because the in vivo microenviroment somehow activates the Wwox apoptotic pathway.
This study demonstrates for the first time that WWOX induces cell growth inhibition and apoptosis in lung cancer cells. In A549 and H460 cell lines we observed caspase-dependent induction of apoptosis through the intrinsic pathway. In H1299 cells we observed cleavage of full length PARP-1 but neither procaspase 3, nor 9, nor 8 were cleaved, possibly because apoptosis occurs later in these cells. Wwox and Fhit protein expression is frequently reduced in lung, breast and bladder cancers in association with promoter hypermethylation (74). Epigenetic alterations can be reversed by specific agents
48 or inhibitors, suggesting such inhibitors as therapeutic agents (75-79). The ponA inducible expression of Wwox can be considered a model for the effects of WWOX reactivation, after silencing by epigenetic mechanisms. The extent of loss of tumorigenicity after restoring inducible Wwox expression was comparable to the tumor suppression observed after Ad-WWOX expression, both in vitro and in vivo, suggesting that massive overexpression of Wwox is not necessary to effect tumor suppression. This suggests that drugs capable of reactivating the epigenetically silenced WWOX gene, could be effective in treatment of lung cancer.
In conclusion, restoration of Wwox protein expression in lung cancer cells is followed by induction of apoptosis in vitro and suppression of tumorigenicity in vivo and suggests that reactivation of the Wwox signal pathway is a potential target for lung cancer prevention and therapy.
Publication: Fabbri M*, Iliopoulos D*, Trapasso F, Aqeilan RI, Cimmino A, Zanesi N, Yendamuri S, Han SY, Amadori D, Huebner K, Croce CM. WWOX gene restoration prevents lung cancer growth in vitro and in vivo. Proc Natl Acad Sci U S A. 102: 15611-6 (2005). *contributed equally
49
CHAPTER 4
EPIGENETIC RESTORATION OF WWOX SUPPRESSES TUMORIGENICITY
IN VITRO AND IN VIVO
4.1 Abstract
Epigenetic changes involved in cancer development, unlike genetic changes, are reversible. DNA methyltransferase and histone deacetylase inhibitors show anti- proliferative effects in vitro, through tumor suppressor reactivation and induction of apoptosis. Such inhibitors have shown activity in the treatment of hematologic disorders but there is little data concerning their effectiveness in treatment of solid tumors. FHIT,
WWOX and other tumor suppressor genes are frequently epigenetically inactivated in lung cancers. Lung cancer cell clones carrying conditional FHIT or WWOX transgenes showed significant suppression of xenograft tumor growth after induction of expression of the FHIT or WWOX transgene, suggesting that treatments to restore endogenous Fhit and Wwox expression in lung cancers would result in decreased tumorigenicity. H1299 lung cancer cells, lacking Fhit, Wwox, p16INK4a and Rassf1a expression due to epigenetic modifications, were used to assess efficacy of epigenetically targeted protocols in suppressing growth of lung tumors, by injection of 5-Aza-2-deoxycytidine (AZA) and
50 Trichostatin A (TSA) in nude mice with established H1299 tumors. High doses of intraperitoneal AZA/TSA suppressed growth of small tumors but did not affect large tumors (200 mm3); lower AZA doses, administered intraperitoneally or intratumorally, suppressed growth of small tumors without apparent toxicity. Responding tumors showed restoration of Fhit, Wwox, p16INKa, Rassf1a expression, low mitotic activity, high apoptotic fraction and activation of caspase 3. These preclinical studies show the therapeutic potential of restoration of tumor suppressor expression through epigenetic modulation and the promise of reexpressed tumor suppressors as markers and effectors of the responses.
4.2 Introduction
Lung cancer is the leading cause of cancer death in the western hemisphere for men and women, eclipsing the combined mortalities from breast, colon, and prostate cancers. Genomic aberrations involving chromosome 3p are the most frequent and earliest genetic events in lung carcinogenesis (73,80). In particular, the FHIT gene, spanning the human common fragile site, FRA3B, at chromosome 3p14.2, and the
RASSF1A gene at 3p21 are frequently silenced in lung cancer and show hallmarks of tumor suppressor genes (81). FHIT is altered in many other types of cancers, including breast, head and neck, cervical, bladder, esophageal, gastric, pancreatic, and renal cancer and alterations occur in very early stages of preneoplasia in some organs. Fhit loss is associated with progression and outcome in cancers of various organs (20). Similarly, the WWOX gene at FRA16D is frequently silenced in lung cancers, and has recently
51 been shown to suppress growth of lung cancer cells in nude mice (48,74,82). The
RASSF1A promoter region is hypermethylated in a large fraction of lung and other cancers (83,84) and methylation of the RASSF1A regulatory region has been proposed as a biomarker of lung cancer (85).
Exogenous Fhit protein expression induces apoptosis and retards tumor cell proliferation in vivo and in vitro (86-88). Re-expression of Fhit in many Fhit-deficient cancer cells suppresses tumorigenicity and FHIT-viral gene therapy prevents and reverses carcinogen-induced gastric cancers in Fhit-deficient mice (87,89-92). Studies confirming the role of FHIT as a suppressor gene have used cancer cells infected with adenoviral-
FHIT or stably transfected with FHIT-expressing plasmids, both of which cause high levels of constitutively expressed Fhit protein. Due to anatomical barriers associated with lung cancer and the need for high multiplicities of adeno or other virus for effective delivery, the therapeutic potential of FHIT gene therapy in lung cancer has not been fully examined in preclinical models. To further investigate this potential, we used hormone- inducible FHIT expressing lung cancer cells in a nude mouse xenograft model to investigate the effect of differing Fhit expression levels on tumorigenesis and the potential for pharmacologically directed gene therapy.
In addition to deletions and translocations, the FHIT gene, as well as the fragile
WWOX gene, is frequently silenced by hypermethylation of its CpG dinucleotide-rich promoter regions (74,93-96). The RASSF1A and CDKN2A genes at chromosome regions 3p21 and 9p21, respectively, encode the tumor suppressors, Rassf1a and p16, that are also frequently silenced in lung cancers (97,98). Because promoter hypermethylation is responsible for silencing FHIT in >30%, WWOX in >40%, RASSF1A in >32%, and
52 CDKN2A in >66% of NSCLCs (74,97,98), pharmacologic therapy to restore expression of Fhit, Wwox and other tumor suppressors, for treatment of NSCLC offers a potentially efficacious approach. We hypothesized that use of demethylating agents in treatment of nude mouse NSCLC xenografts, that lack expression of Fhit, Wwox and other suppressors, due to promoter hypermethylation, would cause re-expression of the silenced tumor suppressors, and result in decreased tumorigenicity. If the hypothesis were confirmed, then xenograft models of various types of tumors, whether localized or systemic, could be used to optimize treatment routes and dosage, as well as to test markers and effectors of tumors suppression, and to explore genome-wide effects of drugs targeted to epigenome remodeling.
4.3 Materials and methods
4.3.1 Cell culture and isolation of conditional H1299 Fhit expressing clones.
The FHIT-deficient human NSCLC H1299 cell line was purchased from American Type
Culture Collection and maintained in MEM supplemented with 10% FBS and 100 µg/ml gentamicin. H1299 cells stably transfected with the ecdysone-inducible vector pVgRXR
(Invitrogen, Grand Island, NY, USA) were obtained from Dr. Jennifer Pietenpol and maintained in medium containing 200 µg/ml Zeocin. These cells were stably transfected with pIND-FHIT, and G418 resistant clones, including H1299D1, were tested for expression of Fhit before and after induction with 1 µM PonA (Invitrogen, Grand Island,
NY, USA). Immunoblot analysis of H1299D1protein lysates showed very low Fhit expression without PonA and a steady increase after induction.
53 4.3.2 In vitro induction and re-expression of Fhit.
Fhit expression was induced in H1299D1 cells by culturing with 1 µM PonA. Because maximal Fhit expression was achieved following 48-72 h of induction, 48 h was used in subsequent studies. DNA demethylation and histone deacetylation was effected in vitro by treating H1299 cells with 5 µM AZA and 1 µM TSA. Maximal Fhit re-expression was observed with the following treatment regimen, which was used in subsequent experiments: 1x106 H1299 cells were plated in a 75 cm2 flask and grown as described above; every 24 h for 5 days, medium was replaced with fresh medium containing 5
µΜ AZA. After the 5th day, medium was replaced with medium containing 1 µM TSA and incubation continued for 24 h and cells collected for use.
4.3.3 DNA isolation, bisulfite modification and methylation-specific PCR.
Genomic DNA was isolated from cells and bisulfite modified as described (88). Modified
DNA was purified using Wizard DNA purification resin (Promega Corp., Madison, WI), denatured in NaOH, ethanol precipitated, resuspended in 20 µl water and stored at -20 oC.
Bisulfite modified genomic DNA (2 µl) was amplified for sequencing, using primers described previously6. Specific primers covering CpG sites were designed to distinguish, by direct sequencing, between methylated and unmethylated sequences of bisulfite- modified DNA. The genomic location of primers was described previously (74).
54 4.3.4 Real-time PCR analysis.
Real-time PCR analysis was based on fluorescent SYBR Green I dye methodology as described previously (74).
4.3.5 Immunoblot Analysis.
Western blot analysis was performed as described previously (91). Rabbit anti-human
Fhit (1:10,000), rabbit anti-human Wwox (1:20,000) (28) and rabbit anti-human cleaved caspase 3 (1:500) (Santa Cruz Inc) polyclonal sera were used.
4.3.6 Animal Studies.
Female Nu/Nu mice 6-8 wk of age were obtained from Charles River Laboratories
(Wilmington, MA, USA). Mice were maintained, and animal experiments conducted, under institutional guidelines established by the Animal Care and Use Facility at Thomas
Jefferson University and the University of Pennsylvania.
4.3.7 Tumorigenicity studies of H1299 cells conditionally expressing Fhit.
Two control groups of 5 mice each were used in tumorigenicity studies. The Fhit negative H1299 and H1299/VgRxR cells were used to establish sc tumors. Briefly, 5x106 cells, either H1299 or H1299/VgRxR, were injected sc into the flank. Tumor diameters were measured 2-3 times per wk and tumor volumes calculated using the equation V
(mm3) = a x b2/2, where a is the largest diameter and b is the perpendicular diameter.
Three groups of 5 mice each were injected with 5x106 stably transfected H1299D1 cells.
The first group of mice received sc flank injections of H1299D1 cells that had been
55 treated, in vitro, with 1 µM Pon A for 48 h. The second group received sc flank injections of H1299D1 cells that were not treated with PonA. The third group received flank injections of H1299D1 cells that were not treated in vitro with the inducer PonA, but were administered 4 weekly in vivo injections of 125 µg PonA in 100 µl DMSO, beginning 2 wk following inoculation of cells. PonA was administered by injection with a
27-gauge needle into the tumor or tumor cell inoculation area. Tumor measurements and volume calculations were performed as described above.
4.3.8 Tumorigenicity studies of H1299 cells: effects of epigenetic therapy.
In vitro treatment: Three groups of 5 mice each were used to determine the effects of in vitro treatment with either AZA or AZA and TSA on tumorigenesis of H1299 cells. The first group received sc flank injections of 7x106 H1299 cells that had been treated, in vitro, with 5 days of AZA. The second group received sc flank injections of 7x106 H1299 cells that had been treated, in vitro, with 5 days of AZA and 1 day of TSA. Control group mice received sc flank injections of 7x106 untreated H1299 cells. In vivo treatment:
Protocol 1: 2 groups of 5 mice each were used to determine the effect that in vivo, systemic administration of AZA and TSA had on existing tumors. All mice received sc flank injections of 7x106 H1299 cells. When mean tumor volumes for all ten mice reached ~175 mm3, in vivo treatment was initiated. Group 1 mice received ip injections of
100 µl PBS on days 24 and 25, and days 31 and 32 post-inoculation. Group 2 mice received ip injections of 100 µl of an AZA/TSA combination (7.5 mg/kg AZA dissolved in 50 µl sterile water plus 700 µg/kg TSA dissolved in DMSO) on days 24 and 25, and days 31 and 32 post-inoculation. Protocol 2: 4 groups of 5 mice were used to optimize 56 the effect of in vivo AZA and TSA treatment on existing tumors. All mice received sc flank injections of 7x106 H1299 cells. When mean tumor volumes reached 50-70 mm3
(day 17), treatment was initiated. In addition to initiating treatment at an earlier time point, the drug dosage was lowered to 5.0 mg/kg AZA and 500 µg/kg TSA. Treatments were administered on days 17 and 18, and days 24 and 25. Group 1 (IP PBS) mice received ip injections of 100 µl PBS, group 2 (IP AZA/TSA) received ip injections of
“low dose” AZA/TSA. Groups 3 (IT PBS) and 4 (IT AZA/TSA) mice received 100 µl
PBS and “low dose” AZA/TSA it, respectively. Protocol 3: 3 groups of 4 mice were used to determine the effects that each drug, given as monotherapy, had on tumor growth and tumor suppressor re-expression. Due to deaths of 2 mice that had received the high- dose AZA/TSA combination therapy (from “protocol 1,” above), we sought to determine if either drug, given singly, was responsible for the observed toxicity. As in protocol 2, treatments were initiated when mean tumor volumes reached 50-70 mm3, and were thus given on days 17 and 18, and days 24 and 25. Group 1 mice received ip injections of 100
µl PBS. Group 2 received ip injections of 7.5 mg/kg AZA, and group 3 received ip injections of 700 µg/kg TSA.
4.3.9 Statistics.
The in vivo results were expressed as mean ± SDs. Student’s two-sided t test was used to compare the values of the test and control samples. A value of P < 0.0001 was considered significant.
57 4.4 Results
4.4.1 In vitro Fhit induction in H1299D1 cells.
H1299D1, non small cell lung cancer-derived cells, are stably transfected with a FHIT cDNA that can be induced by treatment with the insect hormone, ponasterone A
(PonA)24. Fhit is not expressed in H1299 cells while uninduced H1299D1 cells express a low basal level of Fhit and maximal expression is observed after treatment with PonA for
48 h (Figure 4.1A). Flow cytometric analysis of D1 cells uninduced and induced with
PonA for 48 h did not show notable differences in cell cycle kinetics (not shown) and growth curves were also similar in vitro, confirming that effects of exogenous Fhit expression are not evident in vitro24.
4.4.2 Effect of Fhit induction on tumorigenesis.
H1299 cells, as well as H1299/pVgRXR, H1299D1, and H1299D1 induced in vitro, were injected into nude mice subcutaneously (sc), and animals observed for appearance of tumors. Tumors of control group 1 (H1299) reached 400 mm3 mean volume by day 30 and control group 2 (H1299/pVgRXR) tumors reached 400 mm3 mean size by day 32.
Group 3 tumors, inoculated with H1299D1 cells (basal Fhit expression), reached 400 mm3 mean volume by day 61 with one mouse remaining tumor free. Group 4 tumors, of mice inoculated with in vitro transiently induced H1299D1 cells, reached 400 mm3 mean volume by day 70, with one mouse tumor free. Tumors of group 5, in which H1299D1 cells were inoculated and in vivo Fhit induction by intratumoral (it) administration of
PonA begun 2 wk later, reached 400 mm3 mean volume by day 81, with 2 mice remaining tumor free (Figure 4.1B). All tumors at endpoint were negative for Fhit by
58 immunoblot as shown in the inset of Figure 4.1B, suggesting selective growth of cells that had lost or silenced the FHIT plasmid.
4.4.3 Hypermethylation of suppressor gene promoters in H1299 cells.
Methylation specific PCR amplification, using primers specific for methylated or unmethylated DNA sequences, performed on FHIT and WWOX regulatory regions of
H1299 DNA, demonstrated amplification only of methylated sequences (Figure 4.2A).
Direct bisulfite sequencing revealed that FHIT regulatory regions are heavily methylated
(Figure 4.2B). Treatment of H1299 cells with AZA, in vitro for 5 days, resulted in progressive demethylation of FHIT intron 1 and promoter sequences (Figure 4.3A). To demonstrate that de novo methyltransferase (DNMT) activity persists, in vitro AZA inhibition of DNMT was stopped after 6 days, resulting in progressive, gradual re- methylation of FHIT regulatory regions. FHIT mRNA expression level correlated with extent of FHIT intron 1 demethylation, supporting the conclusion that loss of Fhit protein expression in H1299 cells is due to methylation associated epigenetic alterations.
Continuous DNMT inhibition with AZA treatment for 12 days resulted in FHIT demethylation and mRNA re-expression (Figure 4.3B) while histone deacetylase
(HDAC) inhibition with TSA alone, affected neither methylation nor expression (Figure
4.3C). It was reported previously that the regulatory regions of the RASSF1A and
CDKN2A genes are also hypermethylated in H1299 cells (100).
59 Figure 4.1. Effect of Fhit expression on tumorigenicity of lung cancer cells. A. Time course of Fhit induction (50 µg protein/lane) in H1299D1 cells after treatment with 1 µΜ
PonA. Gapdh expression served as loading control. Lane 1, H1299 cells (no expression); lane 2, uninduced (low expression); lane 3, 24 h after PonA addition; lane 4, 48 h after
PonA addition. B. Growth in nude mice of H1299 parental cells and PonA induced
H1299D1 cells. At day 30, tumors in the 5 mice inoculated with H1299 parental cells reached an average of 400 mm3 and were sacrificed. H1299 D1 cells pretreated in vitro (1
µΜ PonA for 48 h) or in vivo (125 µg PonA ip, 2 wk after inoculation) and uninduced
H1299D1 cells, developed tumors at later times, as summarized in the table below the graph. Protein lysates of tumors at endpoints were negative for Fhit expression; see inset: lane 1, MCF7 (positive control); lane 2, H1299; lane 3, H1299D1; lane 4, H1299D1 in vitro pretreated; lane 5, H1299D1 in vivo treated.
60
61
A Hela Kato3 BxPC3 H1299 M U M U M U M U FHIT 74 bp Hela MCF7 H1299 M U M U M U WWOX 347 bp
B FHIT promoter exon 1 intron 1
CpG sites RC48 RC49 Kato3 U2020 LNCaP 5637 BxPC3 T24 HeLa SW780 H1299
Figure 4.2. FHIT and WWOX methylation status in cancer cells. A. FHIT methylation specific PCR (MSP) analysis in HeLa (cervical), Kato III (gastric), RC49
(renal), MCF7 (breast) and H1299 (lung) cancer cells. U, amplification using FHIT regulatory region primers specific for unmethylated or, M, methylated DNA sequences.
FHIT and WWOX regulatory regions in H1299 cells show amplification only of methylated sequences. B. FHIT methylation status determined by direct bisulfite sequencing. Each square represents a CpG site; filled squares represent methylated cytosines. In most cancer cells FHIT is methylated in intron 1; H1299 FHIT regulatory sequences are heavily methylated throughout.
62 Figure 4.3 Effect of AZA and TSA treatment on FHIT methylation status and mRNA expression. A. FHIT methylation and expression status in H1299 cells during a
12 day time course. When cells were treated with AZA for days 1 – 5, FHIT intron 1 sequences are increasingly demethylated. When AZA treatment was stopped on day 6,
FHIT intron 1 was re-methylated and mRNA expression was silenced beginning on day
8. B. Continuous AZA treatment for 12 days resulted in FHIT intron 1 demethylation and mRNA expression. C. TSA treatment did not affect either FHIT DNA methylation levels or mRNA expression.
63
64 4.4.4 Fhit and Wwox protein re-expression in H1299 cells following AZA and TSA treatment.
In vitro treatment of H1299 cells with AZA and AZA followed by TSA, resulted in marked re-expression of Fhit and Wwox proteins, as demonstrated by western blot.
Maximal expression was achieved with 5 days of AZA treatment followed by 1 day of
TSA treatment (Figure 4.4). Densitometry analysis of Fhit and Wwox expression relative to Gapdh demonstrated a >>5-fold increase in Fhit expression (Figure 4.4A) and
>>3-fold increase in Wwox expression (Figure 4.4B) by combined AZA/TSA treatment.
Also it has been reported that AZA treatment reactivates Rassf1a and p16INKa in H1299 cells (100).
4.4.5 Effect of AZA and TSA treatment on tumorigenesis.
Untreated H1299 cells and H1299 cells that had been treated in vitro with AZA or AZA and TSA, were injected subcutaneously into the flanks of nude mice. The in vitro treated cells were non-tumorigenic, with no mice from either group developing tumors (Figure
4.5A). In contrast, all 5 mice from the control group, which had received untreated
H1299 cells and ip injections of PBS, developed tumors (mean volume >500 mm3) by day 31. Systemic treatment with ip injection of 7.5 mg/kg AZA and 700 µg/kg TSA, resulted in varying effects. Tumors >200 mm3 at time of treatment continued to grow with kinetics similar to that of mice treated with PBS. Mice with tumors that were ≤100 mm3 responded to AZA/TSA treatment by halting tumor growth (Figure 4.5B). Two mice with ~100 mm3 tumors died after the second cycle (4th treatment) of in vivo AZA and TSA. Tumors retrieved at endpoint were tested for expression of Fhit, Wwox and
65 cleaved caspase 3 after preparation of protein lysates. The ~100 mm3 tumor from the surviving mouse was positive for Fhit, Wwox and cleaved caspase 3 by western blot but not p16inka and Rassf1a (Figure 4.5E, lane 4). The two large tumors from treated mice, and all tumors from PBS treated mice, were negative for all the tested proteins (Figure
4.5E, lanes 5, 6). In addition, visual inspection of H&E stained tumor sections revealed lower mitotic activity and higher apoptotic fraction in the responding AZA/TSA treated tumors (Figure 4.5F).
Protocols modified to reduce toxicity and further define efficacy were tested next.
Mice were treated earlier (days 17-18 and 24-25) than previously (days 24-25 and 31-32) and with lower doses of AZA (5 mg/kg) and TSA (500 µg/kg), ip or intratumorally (it).
All tumors of mice treated with AZA/TSA ip or it responded to combination therapy
(Figure 4.5C). Intratumoral AZA/TSA administration was more effective than ip administration and resulted in higher expression of Fhit, Wwox and Rassf1a, greater activation of caspase-3 and re-activation of p16inka (Figure 4.5E, lanes 14-15).
To determine if AZA, TSA or the combination was most effective and least toxic in vivo, mice were treated with AZA or TSA alone. None of the mice treated with AZA or TSA alone died, suggesting that the combination of these agents caused the observed mouse deaths. The groups of mice treated with AZA only or with the lower dose of
AZA/TSA showed almost complete suppression of tumor growth (Figure 4.5C,D).
Tumors in mice treated with TSA alone were not suppressed (Figure 4.5D); thus, histone deacetylase inhibition, without promoter demethylation, did not induce tumor suppressor expression. Expression of tumor suppressors p16INK4a and Rassf1a in H1299 lung cancer cells is also not reactivated by TSA (100).
66
A AZA (μΜ) - - 2 5 1 2 TSA (μΜ) - -1 - - 1 1 - - - Fhit
Gapdh Relative 0 0 1 3.1 3.7 4.8 expression
B
AZA (μΜ) - 2 2 5 TSA (μΜ) - 1 1 - - - Wwox Gapdh Relative expression 0 1 2.9 3.1
Figure 4.4. Fhit and Wwox protein re-expression in H1299 cells after AZA and TSA treatment. Protein lysates from H1299 cells with indicated treatments were immunoblotted for Fhit, Wwox and Gapdh. Densitometry analysis of Fhit expression relative to Gapdh levels is shown. Combination of 5 µΜ ΑΖΑ and 1 µΜ TSA induced
Fhit expression >5-fold and Wwox expression >3-fold, relative to Gapdh level.
67 Figure 4.5. In vivo epigenetic therapy of H1299 xenografts. A. Comparison of tumor volumes in mice inoculated with untreated H1299 cells, or H1299 cells pretreated in vitro with 5 µΜ AZA or pretreated in vitro with 5 µΜ AZA and 1 µΜ TSA. B. Tumor volumes in mice inoculated with untreated H1299 cells and given epigenetic therapy after engraftment. Mice received ip injections of PBS or AZA (7.5 mg/kg)/TSA (700 µg/kg) on days 24 and 25 (TX1), days 31 and 32 (TX2). AZA/TSA in vivo 1 and 2 mice had large tumors (>200 mm) at TX1 and treatments did not prevent continued growth.
AZA/TSA in vivo 3, 4 and 5 tumors were smaller (~100 mm) at TX1, and did not increase in size. The plot for the PBS control tumors is an average from tumors of 5 mice. C. Tumor volumes in mice inoculated with PBS treated H1299 cells or given epigenetic therapy after engraftment. One group of mice received ip injections and another group it injections of PBS or AZA (5 mg/kg)/TSA (500 µg/kg) on days 17 and 18
(TX1), days 24 and 25 (TX2). Ip and it AZA/TSA groups showed suppression of tumor growth. The plot for the PBS control tumors is an average of tumor growth in 5 mice. D.
Tumor volumes in mice inoculated with H1299 cells (control group), AZA (7.5 mg/kg) or TSA (700 µg/kg). E. Fhit, Wwox, Rassf1a, p16 and cleaved caspase 3 protein expression in tumors at endpoint. Lanes 1, 2, 3, tumors from 3 mice treated with PBS
(control); lane 4, AZA/TSA vivo 4 (very small tumor); lane 5, AZA/TSA vivo 1 (large tumor); lane 6, AZA/TSA vivo 2 (large tumor); lane 7, ip PBS1; lane 8, ip PBS2; lane 9, ip AZA/TSA1; lane 10, ip AZA/TSA2; lane 11, ip AZA/TSA3; lane 12, it PBS1; lane 13, it PBS2; lane 14, it AZA/TSA1; lane 15, it AZA/TSA2. For each time point the tumor volumes of each group were compared and P values calculated to determine statistically significant differences. The asterisks in the graphs denote the first time point at which P
68 values were less than 0.0001). F. H&E analysis of tumor sections. The Y axis represents average number, for 3 tumors each, of mitotic figures and apoptotic bodies in 10 high power fields (HPF; x400), from H&E slides of ip PBS or AZA/TSA treated tumors.
69
70
4.5 Discussion
Re-expression of Fhit in most Fhit-deficient cancer cells suppresses tumorigenicity and FHIT-viral gene therapy has been shown to prevent and reverse tumor growth in murine cancer models (87,89-91,101). Very recently, it was shown that Fhit-/-
Vhl+/- mice are more susceptible to lung carcinogenesis than either parental recombinant mouse strain, suggesting that Fhit deficiency combined with haploinsufficiency for the
VHL gene, a locus that is frequently hemizygous in lung cancers, creates a permissive environment for lung cancer development; this permissive environment is created in many lung cancer victims by exposure to carcinogens in tobacco smoke or other environmental exposures (102). These and other studies have suggested that replacement of an active FHIT gene in lung and other cancers would have a therapeutic effect.
Utilizing an ecdysone inducible FHIT-gene stably introduced into an endogenous Fhit negative NSCLC cell, we have demonstrated, with proof-of-principle experiments, that pharmacologically induced restoration of a silent FHIT gene is a potentially efficacious approach to treatment of NSCLC.
FHIT induction in H1299D1, following treatment with PonA, either in vitro or in vivo, resulted in a significant decrease in tumorigenesis. While H1299D1 cells express a level of Fhit barely detectable by immunoblot, tumorigenicity of these cells was significantly delayed, requiring twice as long to reach the size of parental H1299 tumors, suggesting that even minimal Fhit expression can suppress lung tumorigenesis. The transiently induced H1299D1 cells, expressing maximal Fhit at inoculation, showed a growth delay of 2.3-fold; ie, these tumors took more than twice as long to reach the size
71 of H1299 parental cells; growth of tumors after inoculation of H1299D1 cells, followed 2 weeks later by sustained in vivo induction of Fhit expression, was delayed 2.7-fold, with
2 of 5 mice remaining tumor free at endpoint. We hypothesized that the reason for tumor growth after Fhit induction was selection of cells that had lost Fhit expression through loss or silencing of the FHIT plasmid. In fact, each of the tumors isolated and tested by immunoblot at endpoint was Fhit negative. These studies supported the hypothesis that systemic drug therapies that result in Fhit re-expression could decrease the tumorigenicity of Fhit-deficient cancers. Recently, we have shown that Wwox restoration in H1299 lung cancer cells by adenovirus or inducible expression in vitro or in vivo suppresses lung cancer growth (82). Re-expression of RASSF1A in lung carcinoma cells also inhibits tumour formation in nude mice (103) and introduction of exogenous CDKN2A into cancer cells restores normal cell cycling (104). By using both methylation specific PCR amplification and bisulfite DNA sequencing we have shown that the FHIT gene is silenced in the human NSCLC cell line, H1299, by hypermethylation, as is the second fragile tumor suppressor gene, WWOX, and others have shown that RASSF1A and
CDKN2A are silenced in these cells by methylation (97,98).
We also demonstrated that Fhit and Wwox can be re-expressed in H1299 cells by in vitro treatment with the hypomethylating agent, AZA, as shown by previously for
Rassf1a and p16 proteins (97,98). To determine the effect of re-expression of the tumor suppressors on tumorigenesis, we challenged nude mice with H1299 cells that had been treated in vitro with either AZA or AZA and TSA. Treatment of H1299 cells with both hypomethylating monotherapy, or combined hypomethylating and histone deacetylase inhibiting therapy, rendered the cells nontumorigenic, even though the tumor suppressor
72 expression was transient. Importantly, while combination in vitro treatment with AZA and TSA caused transient re-expression of Fhit and Wwox, and rendered H1299 cells nontumorigenic, it did not affect in vitro cell growth. The cells appeared healthy, excluded trypan blue, and were metabolically active by MTS assay (data not shown) at the time of tumor challenge. Coupled with the findings from our H1299D1 experiments, that very low levels of Fhit expression significantly decreased lung tumorigenesis, this data suggests that synchronous transient expression of Fhit, Wwox and/or other tumor suppressors at levels approaching those of normal endogenous expression could result in complete elimination of lung tumor cells.
Most preclinical studies and clinical trials have been performed using DNMT and
HDAC inhibitors in myelodysplastic syndrome or leukemias but not in solid tumors (104-
106). The first clinical trials in solid tumors that used these agents were not successful, probably because the very high drug doses had cytotoxic effects (107). To evaluate, in the lung cancer xenograft model, the potential therapeutic efficacy of drugs with epigenetic chromatin remodeling effects, we administered inhibitors of DNA methylation and histone deacetylation, singly or in combination, by ip or it injections, into mice with established H1299 xenograft tumors. Growth of tumors larger than ~200 mm3 continued with kinetics similar to that in mice receiving PBS ip injections, suggesting that the drugs were not effectively reaching many of the cells in these large tumors. This suggests that demethylating agents may not be very effective in treatment of large solid tumors but could be useful in treating patients with local spread to lymph nodes for which large tumors have been removed by surgery. However, ip administration of AZA and TSA, or
AZA alone, to mice with smaller tumors (≤100 mm3) resulted in marked tumor growth
73 inhibition. Western blot analysis of protein lysates from endpoint tumors illustrated that the large tumors that were unaffected by the drugs did not show evidence of reexpression of the four representative tumor suppressors tested, while the small tumors that did not grow after initiating drug treatments, expressed appreciable Fhit, Wwox, Rassf1a and p16 protein, especially after local drug treatment, again suggesting that delivery of drug within the tumor is key to effectiveness. In addition, further evaluation of the small, suppressed tumors revealed low mitotic activity, high apoptotic fraction and activation of caspase 3, suggesting that systemic epigenetic therapy, in cases of small, or possibly disseminated, tumors, can cause significant anti-tumor effects. It is possible that reexpression of any one of these suppressors might have been sufficient to stop growth of the H1299 tumors, as suggested by the H1299D1 conditional Fhit expressor results. This is important because lung cancers will not all exhibit the same tumor suppressor methylation profile but all lung cancers are likely to have at least one of these genes inactivated by hypermethylation.
The death of some treated animals is probably related to systemic administration of the combination of AZA and TSA, based on the following observations: first, among the animals that died during these studies were 2 that received 4 injections of high dose combination drug; second, in an unrelated, unpublished study, 5 of 5 mice died following
6 injections of high dose AZA and TSA; third, 1 combination treated mouse at the lower doses, died after the 4th treatment, while none of the single drug treated mice showed signs of morbidity.
Yamada et al., in 2005 reported that DNA hypomethylation can have opposing effects on intestinal and liver carcinogenesis (108). In addition results from decitabine (5-
74 Aza-2-deoxycytidine) clinical trials in patients with hematopoietic malignancies showed that high doses of this drug are cytotoxic (109). To increase specificity of these agents and decrease global effects, it will be necessary to design protocols that effectively induce endogenous tumor suppressors, while minimizing effects on global gene expression and patient toxicity. Our approach, using successively modified treatment protocols in nude mice, shows the effectiveness of the DNMT inhibitor in lower doses, as targeted therapy rather than as cytotoxic agents (110). A recent study by Issa et al. (109) demonstrated that decitabine is effective, specific and not cytotoxic when used in doses
30-fold lower than the maximum tolerated dose. Considering that we observed tumor suppressor reexpression only in tumors that responded to therapy, assessment of reexpression of tumor suppressor genes that are frequently inactivated by promoter hypermethylation may define useful end-point markers for decitabine clinical trials. In addition to studying toxicities of these drugs, it is important to determine the efficacy of drug combinations in various tumor types. As interest in these agents grows, it is not unreasonable to think that small molecules directed at particular genes could be developed.
Publication: Cantor JP*, Iliopoulos D*, Rao AS, Druck T, Semba S, Han SY, McCorkell KA, Lakshman TV, Collins JE, Wachsberger P, Friedberg JS, Huebner K. Epigenetic modulation of tumor suppressor expression in lung cancer xenografts suppresses tumorigenicity Int J Cancer (in press). *contributed equally.
75
CHAPTER 5
THE ROLE OF WWOX IN HORMONE RESISTANT BREAST CANCER
5.1 Abstract
Tamoxifen is commonly used for treatment of estrogen receptor alpha positive breast cancers, but de novo or acquired tamoxifen resistance occurs frequently. We find that
Wwox protein, which binds and retains Ap2α and γ transcription factors in the cytoplasm, mediates tamoxifen sensitivity; Wwox loss initiates tamoxifen resistance through release of Ap2 factors to the nucleus where Ap2α down-regulates PKA-RIα, a
PKA inhibitory subunit, and Ap2γ up-regulates ErbB2 expression. Restoration of Wwox in tamoxifen resistant breast cancer-derived cells restored tamoxifen sensitivity and abrogated ErbB2 expression. Wwox expression is significantly (p<0.01) reduced in tamoxifen resistant breast cancers, confirming results of the in vitro studies. Thus, the
Wwox signaling pathway may provide new targets for therapeutic intervention in antiestrogen-resistant breast cancers.
5.2 Introduction
Breast cancer is the most frequent malignancy among women in western countries (111), although mortality rates have decreased recently due to screening and implementation of adjuvant therapy with tamoxifen, the most commonly used treatment for estrogen
76 receptor alpha (ERα)-positive breast cancers (111-114). Although many patients benefit from tamoxifen in the adjuvant and metastatic settings, half of the breast tumors that recur after therapy are resistant to tamoxifen (115,116). Despite many studies of breast cancers and derived cell lines with acquired or selected resistance to tamoxifen treatment, the mechanisms involved in resistance are not fully understood (115-117).
Several studies have suggested an association between over expression and aberrant activity of ErbB2 and tamoxifen resistance in human breast cancers (118,119).
Also SRC-3 (AIB1) up-regulation and stabilization of the interaction between ERα and
SRC-1 by Cyclin D1 have been related to tamoxifen resistance in vitro (120-122).
Recently, a correlation was reported between down-regulation of the inhibitory subunit of protein kinase A (PKA-RIα) and tamoxifen resistance (123). Activation of PKA by PKA-
RIα down-regulation leads to phosphorylation of ERα at serine 305, converting tamoxifen from an ERα inhibitor to a growth stimulator. The mechanisms by which
PKA-RIα is down-regulated and ErbB2 up-regulated in tamoxifen resistant cases were unknown.
We noted that Wwox protein expression was reduced in a large fraction of breast cancers (28), and nearly lost in a subclone of MCF7 cells selected for tamoxifen resistance in vitro. We have also shown recently that Wwox is down-regulated in breast cancers due to DNA hypermethylation in its regulatory region (74). WWOX, a gene with tumor suppressor hallmarks, encompasses common fragile site, FRA16D, in a chromosome region involved in allelic loss in breast cancers (23). The gene encodes a 46- kDa protein containing two WW domains that play an important role in Wwox function
(23, 42-44). Like the SH3 domain, the WW domain is characterized by interaction with 77 proline-containing ligands and mediates protein-protein interaction (38,39). The Wwox
WW domains were predicted to interact with several proteins of interest in breast cancer, including p73, the cytoplasmic domain of ErbB4 and the Ap2 transcription factors, using the ProChartTM database (Cytogen Corporation, Princeton, NJ), the largest repository of protein-protein interaction data for the WW domain in the human proteome, and confirmed through biological studies (42-44,124). We report that breast cancer cells in which Wwox expression was knocked down by different RNAi methods, such as siRNA and lentivirus-based shRNA, lost sensitivity to tamoxifen-mediated growth inhibition.
We hypothesized that Wwox mediates tamoxifen sensitivity through its cytoplasmic interaction with transcription factors, and examined in detail the Wwox signal network that controls tamoxifen sensitivity.
5.3 Materials and methods
5.3.1 Antisera and drugs.
The following antisera and reagents were used: anti-Wwox (1:20,000) (28); anti-Gapdh
(1:10,000) (Santa Cruz); anti-Ap2γ (1:500) (6E4/4, Santa Cruz); anti-Ap2α (1:250) (3B5,
Santa Cruz); anti-ERα (1:500) (HC-20, Santa Cruz); anti-ErbB2 (1:200) (C18, Santa
Cruz); anti-PKA-RIα (1:250, BD Biosciences); anti-phospho-ErbB2 (Tyr1248) (1:200)
(2247S, Cell Signaling). Mouse monoclonal anti-ErbB2 (clone e2-4001+3B5) at 1:500 diluted was used (LabVision/Neomarkers, Fremont, CA). 4-hydroxytamoxifen was purchased from Sigma and ICI 182,780 (Fulvestrant) from Tocris Cookson Ltd
(Ellisville, MO).
78
5.3.2 Cell culture.
MCF7, BT-474, T47D, MDA-MB-231, MDA-MB-453, HCC1937 breast cancer-derived cell lines were obtained from American Type Culture Collection (ATCC) and maintained in Dulbecco Modified Eagle Medium (DMEM) (Sigma) with 10% FBS and 2 mM L- glutamine (Cellgro). Before drug treatments cells were maintained for 2 days in phenol- red free DMEM (Sigma) supplemented with 10% charcoal/dextran-treated FBS
(BioGemini) and 2 mM L-glutamine.
5.3.3 Establishment and characterization of the MCF7 cell line with acquired tamoxifen resistance.
MCF7 cells resistant to tamoxifen (MCF-OHTr cells) were maintained in phenol-red free
DMEM with 10% charcoal/dextran-treated FBS, 2 mM L-glutamine and 100 nM 4- hydroxytamoxifen. We used a single estrogen-responsive MCF7 clone stably transfected with an ERα-responsive luciferase reporter (ERE-pS2-Luc) (125) to derive a subline resistant to tamoxifen (MCF7-OHTr). The stably integrated ERE-pS2-Luc reporter was used to monitor ERα transcriptional activity. The MCF7-OHTr subline was obtained by growing MCF7 cells (1x106) in hormone-free medium containing 10-7 M 4- hydroxytamoxifen (OHT; the active metabolite of tamoxifen). In the presence of OHT,
MCF7 cells first underwent growth arrest (weeks 1-2), followed by a period of modest cell death (weeks 2-8) and slowed cell proliferation (weeks 4-8), and eventually achieved a steady growth rate, with a weekly split ratio of 1:5, after four months of OHT treatment.
During the first eight weeks of OHT treatment, a layer of the cells remained evenly
79 distributed on the culture surface, despite the observed growth arrest and cell death, suggesting adaptation of a significant number of MCF7 cells to OHT-treatment. ERα transcriptional activity in MCF7-OHTr cells was examined by monitoring the expression levels of the stably integrated ERE-pS2-Luc. Compared to MCF7, basal luciferase activity was higher in MCF7-OHTr (~2-fold) (data not shown). Cotreatment with tamoxifen decreased estrogen-induced luciferase expression in both MCF7 and MCF7-
OHTr cells. These observations suggest that ERα retains its transcriptional activity and sensitivity to different ligands in MCF7-OHTr cells.
5.3.4 RNA interference methods. i) small interference RNA (siRNA): WWOX siRNA and negative control siRNA were purchased from Ambion, Inc (Austin, TX). 4x106 cells grown in phenol-red free DMEM supplemented with 10% charcoal/dextran-treated FBS and 2 mM L-glutamine were transfected with WWOX siRNA or siRNA negative control using siPORT NeoFx lipid- based transfection agent (Ambion, Inc) in six-well plates. 24 h later cells were treated with 1 µM 4-hydroxytamoxifen and used in analyses. ii) lentivirus-delivered short hairpin
RNA against WWOX: Sense (5’-ccgggccaagaatgtgcctcttcatctcgagatgaagaggcacattcttgg ctttttg-3’) and antisense (5’-aattcaaaaagccaagaatgtgcctcttcatctcgagatgaagaggcacat tcttggc-
3’) oligos were annealed and ligated in AgeI and EcoRI sites of pLKOpuro.1 retroviral vector (126). WWOX shRNA lentivirus was produced by co-transfection with PCMV-
VSV-G and pHR-CMV vectors, using the calcium phosphate method in 293T cells for 2 days and virus-containing supernatant was removed by centrifugation. Breast cancer cells were infected with lentivirus-shRNA against WWOX (shWWOX) for 48 h followed by
80 tamoxifen treatment for 4 days. iii) WWOX shRNA expression plasmid: a designed microRNA against WWOX 3’UTR was cloned into pSHAG-MAGIC2 (pSM2) (Open
Biosystems, Inc). Breast cancer cells were transiently transfected with this plasmid DNA using Arrest-In tranfection reagent for 48 h (Open Biosystems Inc). WWOX knock-down stable MCF7 clones were isolated after transfection with this WWOX shRNA plasmid, and selection for 3 wk in 0.5 µg/ml puromycin. Clones were tested for Wwox down- regulation by immunoblot followed by densitometry analysis (QuantImage 5.2 program).
5.3.5 Cell proliferation and death assays. i) Cell counts. 3-5x105 breast cancer cells were plated in 6-well plates and cell suspensions from treated and untreated cells were counted in a Beckman Coulter Counter using vi-Cell viability program (analyzing 100 images). Cell growth was calculated as a percentage relative to untreated counterpart. To determine significant differences in cell growth, student’s t test was used and p <0.05 was taken as significant. ii) Flow
Cytometry Analysis. Cells were fixed in ethanol, collected by centrifugation, stained with propidium iodide and analysed by flow cytometry by EPICS-XL scanner (Beckman-
Coulter). iii) CCK-8 cell toxicity assay. CCK-8 is a sensitive nonradioactive colorimetric assay for determining cell growth (Dojindo Molecular Technologies, Japan). The inhibition factor was calculated as fold difference relative to untreated counterpart cells;
96 well plates were used for this experiment and each cell line was plated in 12 wells.
5.3.6 Immunoblotting and immunoprecipitation.
81 Protein extraction and immunoblot analysis was performed as described previously (42).
Protein-protein interactions were detected by co-immunoprecipitation followed by imunoblot analysis. Briefly cells were lysed in buffer containing 50 mM Tris (pH 7.5),
150 mM NaCl, 10% glycerol, 1% NP40 and protease inhibitors. Lysates were precleared with IgG for 1 h and immunoprecipitations were carried out overnight in the same buffer using 500 µg of protein. Precipitates were washed 4 times with the same buffer containing 1% NP40 and 0.1% SDS and assessed by immunoblot analysis.
5.3.7 Immunofluorescence.
Subcellular localization of Ap2γ (1:200), Ap2α (1:200) and Wwox (1:5,000) proteins was determined by immunofluorescence using specific antisera. Cells were seeded on fibronectin-coated cell culture slides (Becton Dickinson), fixed for 10 min in 3.7% PBS- buffered formaldehyde, blocked with 3% BSA, and incubated with primary antisera for
Ap2γ (1:200 dilution), Ap2α (1:200 dilution) or Wwox (1:5,000 dilution) for 1 h in 1%
BSA in PBS, and with secondary antisera under the same conditions. Secondary sera used were anti-rabbit FITC-conjugated anti-IgG and anti-mouse Texas red-conjugated anti-IgG (Jackson Immuno Research Molecular Probes).
5.3.8 Luciferase assay.
Transient transfections were done as previously described (Aqeilan et al., 2004), using
FuGENE 6, according to the instructions of the manufacturer (Roche Applied Science,
IN). Cells were transfected with 0.5 µg of the firefly luciferase reporter plasmid and 0.05
µg of the control plasmid containing Renilla luciferase, pRL-TK (Promega). Cells were 82 lysed with passive lysis buffer and Firefly and Renilla luciferase activities were measured consecutively by using dual-luciferase assays (Promega) 24 h after transfection. The total amount of DNA used in the transfections was kept constant by adding a parental vector.
Each transfection was carried out in triplicate in 12-well plates.
5.3.9 Protein Kinase A assay.
Protein kinase A activity was measured using a non-radioactive protein kinase A assay kit (Calbiochem) in a 96-well plate as described (127). This assay is based on an enzyme- linked immunoadsorbent assay (ELISA) that utilizes a synthetic PKA pseudosubstrate and a monoclonal antibody that recognizes the phosphorylated form of the peptide. Each sample was plated in quadruplicate and each experiment was repeated thrice.
5.3.10 Plasmid construction.
The construction of pCMV-Myc-WWOX (wild-type WWOX), pCMV-Myc-WWY33R
(mutant WWOX), Ap2γ/pcDNA3.1, Ap2α/pcDNA3.1 expression vectors have been described previously (43). ErbB2-pGL3-basic vector was kindly provided by Dr. Helen
C. Hurst (University of London, England). PKA-RIα-pGL3 plasmid was constructed by cloning of the PKA-RIα promoter from PKA-RIα-pCAT vector (provided kindly from
Dr. Kjetil Taskén, University of Oslo, Norway) into the XbaI site of pBS-SK vector and then digestion with KpnI/SacI of the PKA-RIα promoter and cloning into pGL3-basic vector (Promega).
83 5.3.11 DNA methylation studies.
MCF7-OHTr cells cells were treated with varying concentrations of AZA (Sigma) and
Wwox re-expression was tested by immunoblot analysis. Gapdh was used as a loading control. Relative expression levels were calculated by densitometry analysis.
5.3.12 Adenoviral WWOX infections.
The WWOX adenovirus was constructed using the Adenovator-CMV5(CuO)-IRES-GFP transfer vector (Qbiogene) as described previously (82). Titers were determined by absorbance measurement (number of viral particles per ml) and plaque assay (plaque- forming units/ml), and transgene expression was assessed by immunoblot using Wwox polyclonal antiserum (28). 5x105 MCF7-OHTr cells were infected (MOI 5) in a 6-well plate, medium changed 24 h post infection, 1 µM tamoxifen was added and cell growth assay performed.
5.3.13 Immunohistochemical analysis.
Fixed tissues of tamoxifen sensitive and resistant breast cancers were stained for Wwox and ErbB2 by immunohistochemistry. The method of staining and features of Wwox and
ErbB2 antisera were described previously (28,30,128). Cases without tumor recurrence in the year following termination of tamoxifen treatment were considered tamoxifen sensitive, whereas cases with cancer recurrences during or in the year following termination of tamoxifen treatment were considered resistant to tamoxifen treatment. One section from each tumor was stained and evaluated for extent and intensity of staining.
For each section a number of fields were scored (>1000 cells/slide). Because of the
84 heterogeneous staining of some tumors, both staining intensity and extent of staining in neoplastic cells were taken into account in expression level scoring. Intensity was graded as strong or reduced expression. Extent of staining was classified as the fraction of stained neoplastic cells: >25%, 25-50%, and >50%. Cases with strong expression in
>50% of the neoplastic cells, were scored as high expression; cases with reduced staining intensity but present in >50% of tumor cells or strong staining in 25-50% of neoplastic cells, were scored as reduced expression; cases with reduced expression in 50% of neoplastic cells or strong staining in <25%, were scored as highly reduced expression.
Also according to the Chi-Square test power analysis we found that a sample size of 96 achieves 84% power to detect an effect size (W) of 0.3000 using 1 degree of freedom with a significance level of 0.05. Calculations were done using SSPS statistical software version 10 (SPSS, Chicago, IL).
5.3.14 Statistics.
Results were expressed as mean ± SDs. Student’s two-sided t test was used to compare the values of the test and control samples. A value of P < 0.05 was considered significant.
For multiple group comparisons ANOVA analysis was performed and to study correlations between the expression levels we calculated the correlation coefficient (R).
5.4 Results
5.4.1 Wwox down-regulation abrogates tamoxifen response
We explored the relationship between Wwox protein levels and tamoxifen response, using three different RNAi approaches (small interference RNA oligos, short hairpin
85 RNA plasmid, lentivirus-based short hairpin RNA) to down-regulate Wwox expression in
MCF7 breast cancer cells. As shown in Figure 5.1A, Wwox expression was reduced by
80% 24 h after transient WWOX siRNA treatment (50 nM), remained reduced for 2 days and began to be re-expressed after 96 h. Even at 120 h after WWOX siRNA treatment,
Wwox protein levels remained very low. Wwox down-regulation abrogated the MCF7 tamoxifen response in comparison to cells treated with scrambled WWOX siRNA
(Figure 5.1B); MCF7-OHTr cells, selected for tamoxifen resistance, showed increased cell proliferation on tamoxifen treatment and Wwox positive MCF7 cells showed greater
(p<0.05) sensitivity to tamoxifen than BT-474 breast cancer cells with reduced Wwox expression. To verify the correlation between Wwox expression and tamoxifen sensitivity, a cell toxicity assay was used to assess tamoxifen effectiveness. Growth of siWWOX treated MCF7 cells was not inhibited by tamoxifen treatment, while tamoxifen was partially effective in BT-474 and T47D cells, which express 16-fold less Wwox than untreated MCF7 cells (Figure 5.1C), confirming the correlation between Wwox level and tamoxifen response. It is known that BT-474 breast cancer cells, which have very high ErbB2 levels are not as sensitive as MCF7 cells to tamoxifen, although they are
ERα positive (129). Lentiviral-WWOX-shRNA infection of MCF7 cells allowed complete Wwox down-regulation (Figure 5.1D) and abrogated tamoxifen response after
48 and 96 h (Figure 5.1E), and transfection with WWOX shRNA plasmid showed similar effects (Figure 5.2). The correlation between Wwox levels and tamoxifen response was observed even at a 10-fold lower tamoxifen dose (100 nM) (Figure 5.1F).
86 87
Figure 5.2. Wwox shRNA plasmid transfection abrogates tamoxifen response. (a)
Wwox is highly down-regulated 48 h after shRNA transfection. (b) Wwox down- regulation by shRNA plasmid inhibits tamoxifen induced apoptosis. (c) In a cell cytotoxicity assay, Wwox down-regulation by WWOX-shRNA plasmid abrogates tamoxifen response.
88 5.4.2 Wwox regulates ErbB2 and PKA signaling pathways.
To explore the mechanism of Wwox-mediated tamoxifen sensitivity, we considered roles for putative Wwox interacting proteins referenced in the ProChart database. ProChart utilized a biochemical approach integrating parallel synthesis of peptides, protein expression and high throughput screening (HTS), combined with bioinformatics to identify Wwox protein interactors (124) (Figure 5.3A). Moderate affinity interaction for
Wwox-Ap2α and high affinity interaction for Wwox-Ap2γ protein was predicted (Figure
5.3B), involving the Wwox WW1 domain and proline-rich motifs in the activation domains of Ap2γ and Ap2α proteins (Figure 5.3C).
To confirm and validate these predictions biologically, we immunoprecipitated endogenous Ap2γ and Ap2α from MCF7 cells and found that both coprecipitate with endogenous Wwox protein (Figure 5.4A). Also endogenous Wwox immunoprecipitation and immunoblot analysis detecting Ap2γ and Ap2α proteins confirmed these protein- protein interactions (data not shown). Aqeilan et al. (43) had demonstrated interaction of overexpressed Wwox and Ap2γ in 293 cells. Bosher et al, showed that Ap2γ binds the
ERBB2 promoter and activates ErbB2 expression in breast cancer cells (130-131), and
Ap2γ and ErbB2 expression are positively correlated in breast cancers (132). Ap2α is overexpressed in breast cancers and binds the promoter of PKA-RIα (133), a regulatory subunit of PKA that is down-regulated in tamoxifen resistant cells and cancers (123). We assessed the levels of Wwox, Ap2γ, Ap2α, ErbB2 and PKA-RIα proteins in a panel of breast cancer-derived cells and observed an inverse relationship between Wwox and
ErbB2 (Figure 5.4B): low Wwox expression levels were associated with high levels of
ErbB2 expression, and very high Wwox expression was associated with very low ErbB2 89 and high PKA-RIα (Figure 5.4B); For example, BT-474 and T47D breast cancer cells exhibit low Wwox levels, high ErbB2 and low PKA-RIα levels; no correlation was observed between Ap2γ, Ap2α and Wwox expression levels.
To determine the basis of these correlations, we examined Wwox, ERα, Ap2γ, ErbB2,
Ap2α and PKA-RIα protein levels at 48 h after WWOX siRNA (50 nM) treatment
(Figure 5.4C). Wwox knock down did not affect ERα, Ap2γ or Ap2α levels, but led to up-regulated ErbB2 and down-regulated PKA-RIα expression. In tamoxifen resistant
MCF7-OHTr cells, Wwox expression was very low, ErbB2 expression was elevated and
PKA-RIα was low, confirming the relationship between Wwox level and ErbB2 and
PKA-RΙα expression. To determine if Wwox controls ErbB2 and PKA-RIα levels through Ap2γ and Ap2α, respectively, ErbB2 and PKA-RIα luciferase reporter constructs were cotransfected with Wwox or Wwox mutant expression plasmid; the mutant Wwox is incapable of interaction with the Ap2 factors (43). In Wwox negative MDA-MB-231 cells, coexpression of Ap2γ with the ErbB2 promoted reporter plasmid led to increased luciferase activity, but in the presence of Wwox co-transfecting plasmid, Ap2γ did not activate the ErbB2 promoted luciferase (Figure 5.4D). Co-transfection with WWOX
WW1 domain mutant (43) partially abolished this effect, indicating that Wwox-Ap2γ interaction regulates ErbB2 levels. To determine if ErbB2 is not only up-regulated but activated, we assessed phosphorylation of ErbB2 tyrosine 1248 in tamoxifen resistant cells and siWWOX transfected MCF7 cells; as shown in Figure 5.4E, ErbB2 was activated by phosphorylation in the tamoxifen resistant MCF7-OHTr and in MCF7 cells with Wwox knocked down by siWWOX.
90
91 Figure 5.4 Wwox mediation of ErbB2 and PKA activity. A: Endogenous Wwox protein interacts with Ap2γ and Ap2α proteins in MCF7 cells. Immunoprecipatation with IgG served as control. B: Relationship of Wwox, Ap2γ, ErbB2, Ap2α, PKA-RIα protein expression levels in breast cancer cells. There is an apparent inverse relationship between Wwox and ErbB2 expression levels and a direct relationship with PKA-RIα levels. C: Wwox, ERα, Ap2γ, ErbB2, Ap2α and PKA-RIα protein levels after 48 h
WWOX siRNA (50 nM) treatment in MCF7 cells. Protein were quantified by densitometry and normalized to Gapdh expression levels. D: Wwox overexpression suppresses ErbB2 activity as shown by luciferase assay. Wwox negative MDA-MB-231 breast cancer cells were transiently cotransfected with ErbB2-Luc (0.2 µg) and Ap2γ (0.5
µg), pCMV-Myc-WWOX expression vector (0.5 µg) or pCMV-Myc-WWY33R vector
(0.5 µg). E: ErbB2 activation after WWOX down-modulation in MCF7 cells and MCF7-
OHTr cells (24 h). 500 µg of cell extracts were immunoprecipitated with anti-ErbB2, followed by immunoblotting and detection with ErbB2 pTyr1248 specific antibody. F:
Wwox overexpression leads to enhanced PKA-RIα transcriptional activity as shown by luciferase assay. MDA-MB-231 breast cancer cells were transiently cotransfected with
PKA-RIα-Luc (0.2 µg) and Ap2α (0.5 µg), pCMV-Myc-WWOX expression vector (0.5
µg) or pCMV-Myc-WWY33R vector (0.5 µg). G: Protein kinase A activity in MCF7 cells transfected with 50 nM WWOX siRNA and in MCF7-OHTr cells. Wwox down- modulation by siRNA in MCF7 cells, or by low endogenous Wwox level in MCF7-OHTr cells led to 7-fold and 5-fold enhanced PKA activity, respectively.
92
93
Results of luciferase reporter assays illustrated Ap2α repression of PKA-RIα expression, while WWOX-AP2α co-transfection restored PKA-RIα expression (Figure
5.4F). As discussed above, low level PKA-RIα expression leads to increased PKA activity (123). To determine if the level of Wwox protein affects PKA activity through control of accessibility of the Ap2α transcription factor to the PKA-RΙα gene, a PKA assay was performed in tamoxifen resistant and siWWOX treated MCF7 cells (Figure
5.4G). Wwox down-regulation in these cells increased the PKA activity 5-7 fold. The results suggest that Wwox, through interaction with Ap2γ, regulates ErbB2 level and activity, and through interaction with Ap2α, regulates PKA-RIα expression and consequently PKA activity. Thus, loss of Wwox allows Ap2γ to activate ErbB2 and
Ap2α to suppress PKA-RIα levels.
5.4.3 Restoration of tamoxifen effectiveness by Wwox up-regulation.
To examine the potential of Wwox up-regulation to restore tamoxifen sensitivity, the tamoxifen resistant MCF7-OHTr cells were transfected with wild type and mutant Wwox expression plasmids (Figure 5.5A). Restoration of Wwox expression restored tamoxifen sensitivity to near parental MCF7 level (Figure 5.5B,C). Mutant WWOX (Y33R) expression did not restore tamoxifen sensitivity, suggesting that Wwox-Ap2 interactions are required for restoration of sensitivity. Restoration of wild type Wwox expression in the resistant cells restored the ability of the cells to undergo apoptosis following tamoxifen treatment for 4 days (Figure 5.5D). Restoration of wild type but not mutant
Wwox expression in MCF7-OHTr cells decreased ErbB2 levels and up-regulated PKA- 94 RIα levels (Figure 5.5E). Also, expression of exogenous Wwox in BT-474 cells, that express high ErbB2 levels, maximized their sensitivity to tamoxifen and suppressed
ErbB2 activity (Figure 5.6). Finally, restoration of Wwox expression by AdenoWWOX virus infection of MCF7-OHTr restored tamoxifen sensitivity (Figure 5.7).
How might Wwox-Ap2α/γ interactions modulate expression of PKA-RIα and
ErbB2? Since Wwox is cytoplasmic (23,43,44) and the transcription factors perform functions in the nucleus, we examined the subcellular location of Wwox and Ap2 proteins in MCF7-OHTr cells by indirect immunofluorescence. Exogenous Wwox was localized in the cytoplasm of MCF7-OHTr cells (Figure 5.8). In untransfected MCF7-
OHTr cells, endogenous Ap2γ and Ap2α proteins were in the nucleus (Figure 5.5F, control panels), where they can activate or repress target gene expression. Expression of exogenous Wwox protein, after transfection with 1 µg of plasmid, allowed sequestration of Ap2γ and Ap2α in the cytoplasm (Figure 5.5F, 1 µg panel), abrogating their transcriptional activity. A lower exogenous Wwox level (0.3 µg) was sufficient to sequester Ap2γ but not Ap2α in the cytoplasm, illustrating the higher binding affinity between Wwox and Ap2γ protein (described in Figure 5.3). To study the effect of the lower binding affinity for Ap2α, we performed PKA activity assays in MCF7-OHTr cells transfected with wild type (0.3 or 1 µg) or mutant WWOX (1 µg) plasmids (Figure
5.5G). Low level Wwox expression (0.3 µg) allowed PKA activation, due to nuclear localization of Ap2α and repression of PKA-RIα, while high level Wwox expression (1
µg) led to low PKA activity (Figure 5.5G), in accordance with the demonstration that
Ap2α is sequestered in the cytoplasm after transfection with 1 µg of Wwox DNA
95 (Figure 3F, 1 µg). Mutant Wwox (1 µg) did not sequester Ap2 proteins (Figure 3F, bottom panels) and did not cause repression of PKA-RΙα, nor inhibition of PKA activity, confirming that tyrosine 33 is important for Wwox-Ap2 protein interactions. The results suggest that Wwox regulates the transcriptional activity of Ap2 factors by sequestration in the cytoplasm, preventing activation or repression of their target genes.
5.4.4 Epigenetic Wwox re-activation sensitizes tamoxifen resistant cells.
We have reported that Wwox expression is epigenetically regulated in breast cancer (74).
Specifically, the WWOX promoter and exon 1 CpG islands were hypermethylated in breast cancers but not in normal mammary tissues (Figure 5.9). Since Wwox expression was low in MCF7-OHTr cells and DNA hypermethylation is a common mechanism of
Wwox inactivation in breast cancer, we wondered if Wwox silencing in the resistant cells might be due to regulatory region hypermethylation. We found that Wwox was re- expressed after treatment of MCF7-OHTr cells with 5-Aza-2-deoxycytidine (AZA), and the levels of up-regulation were dose-dependent (Figure 5.9A); even very low levels of
AZA (0.5 µM) could up-regulate Wwox expression. To study the kinetics of Wwox up- regulation by AZA we treated MCF7-OHTr cells with AZA, for 4 days and then stopped the treatment. Wwox expression was up-regulated after the second day of treatment
(Figure 5.9B) and was still up 4 days (Day 8) after ending AZA treatment (Figure 5.9B).
Cell growth and toxicity assays revealed that up-regulation of Wwox expression could restore tamoxifen sensitivity to the resistant cells (Figure 5.9C, D). Specifically, tamoxifen treatment led to 50% cell growth inhibition of MCF7-OHTr cells (treated with
0.5 µM AZA). 96 Figure 5.5 Restoration of tamoxifen sensitivity by exogenous Wwox expression.
MCF7-OHTr cells were untransfected (CT), transfected with pCMV-Myc-WWOX (WT) or pCMV-Myc-WWY33R (Y33R) plasmids. A: Wwox expression was increased ~5-fold after transfections (quantified by densitometry analysis and listed below the blot). B:
Exogenous Wwox expression restores tamoxifen sensitivity to near parental MCF7 level.
C: Restoration of wild type Wwox leads to cell growth inhibition after 1 µM tamoxifen treatment for 4 days, assayed by a cell cytotoxicity assay. D: Flow cytometry analysis of tamoxifen treated cells (1 µM) before and after Wwox plasmid transfections. E: Wwox restoration leads to down-regulation of ErbB2 levels and up-regulation of PKA-RIα protein levels. F: Fluorescent microscopy of Ap2γ and Ap2α proteins in untransfected, pCMV-Wwox transfected (0.3, 1 µg) and pCMV-WWY33R (1 µg) transfected MCF7-
OHTr cells. A low level of Wwox (0.3 µg) causes redistribution of Ap2γ in the cytoplasm but apparently does not cause cytoplasmic localization of Ap2α. A high (1 µg) Wwox level causes redistribution of both Ap2 proteins in the cytoplasm. G: Protein kinase A activity is reduced in MCF7-OHTr cells transfected with 1 µg Wwox.
97
98
Figure 5.6 Wwox re-expression in BT-474 cells restores tamoxifen sensitivity. A)
Wwox expression levels in untransfected (CT), transfected with wild-type Wwox (WT)
(0.5 µg) or transfected with mutant Wwox (Y33R) (0.5 µg) determined by immunoblot analysis and quantified by densitometry. B) Wild type Wwox up-regulation maximizes tamoxifen response. C) Wild type Wwox up-regulation decreases ErbB2 transcriptional activity measured by luciferase assay.
99
Figure 5.7 Adenoviral Wwox expression restores tamoxifen sensitivity in MCF7-
OHTr cells. A) Time course of Wwox expression levels after Ad-Wwox infection with
MOI 5. B-C) Ad-WWOX infection restores tamoxifen (1 µM) effect 4 days after treatment measured by cell growth assays.
100
Wwox DAPI Merge
Figure 5.8 Wwox subcellular localization in MCF7-OHTr cells after WWOX plasmid transfection. Wwox localizes in the cytoplasm of MCF7-OHTr cells
101
Figure 5.9 Epigenetic Wwox re-activation sensitizes MCF7-OHTr cells to tamoxifen.
A: Dose dependent Wwox re-expression after AZA treatment for 4 days. Even low levels of AZA (0.5 µM) restore Wwox expression. B: Wwox is re-expressed after 2 days of 0.5
µM AZA treatment. Although AZA treatment was stopped at day 4, Wwox was still expressed at day 8. C: Cell viability analysis after 4 days of AZA (0.5 µM) and tamoxifen (1 µM) treatment. AZA treatment restores tamoxifen sensitivity (50% cell growth inhibition). D: Cell toxicity CCK-8 assay after combination treatment (AZA-
TAM). AZA sensitizes MCF7-OHTr cells to tamoxifen (6-fold).
102
5.4.5 Quantitative relationship between Wwox and tamoxifen resistance.
Our in vitro studies have shown that MCF7 cells, with the highest Wwox level, are most sensitive to tamoxifen treatment, while T47D and BT-474 breast cancer cells with reduced Wwox, although ERα+, are insensitive to tamoxifen. In addition, MCF7 cells with Wwox knocked down by siRNA (50 nM) became insensitive to tamoxifen, suggesting a quantitative relationship between Wwox expression levels and tamoxifen response (Figure 5.10A). To further investigate this relationship, we transfected MCF7 cells with various concentrations of WWOX siRNA oligos and down-modulated Wwox protein levels by 35-100% (quantified by densitometry, Figure 5.10B). Down- modulation of Wwox by 35-63% partially abrogated tamoxifen sensitivity, while down- modulation by >65% significantly reduced tamoxifen sensitivity (Figure 5.10B). Down- modulation by ~85% (si40 nM) allowed activation of PKA and ErbB2 signaling pathways (Figure 5.10C). In stable WWOX shRNA MCF7 clones, Wwox down- modulation by about 70% (clone W2A) was critical for activating these pathways
(Figure 5.11). By summarizing results of the 21 different assays described, we quantified the correlation of Wwox, PKA-RIα and ErbB2 protein levels with tamoxifen sensitivity.
Wwox (R2=0.9491) and PKA-RIα (R2=0.8728) levels were directly related to tamoxifen sensitivity, while ErbB2 (R2=-0.7902) level was inversely related to tamoxifen sensitivity
(Figure 5.10D); ie., low Wwox levels led to repression of PKA-RIα, activation of PKA and resistance to tamoxifen, while very low Wwox levels allowed activation of ErbB2 expression. Thus, for breast cancers with very low Wwox expression, alternative therapy that would be unaffected by Wwox levels may be indicated. We have observed that 103 Wwox levels do not affect the ability of ICI 182,780 (fulvestrant) to kill breast cancer cells (Figure 5E) and it is known that most tamoxifen resistant breast cancers respond favorably to treatment with fulvestrant (Michalides et al., 2004; Osborne et al., 2004).
5.4.6 Wwox protein levels and tamoxifen sensitivity in vivo.
To examine the in vivo relevance of the correlation of Wwox expression with tamoxifen sensitivity, we scored the Wwox expression level by immunohistochemical analyses in
53 tamoxifen sensitive and 42 resistant breast cancers. Tamoxifen resistant cases showed reduced Wwox expression (p<0.01) (Figure 5.12); 82% of cases (18 cases) with strong
Wwox expression were tamoxifen sensitive and only 18 % (4 cases) with high Wwox score were resistant. These results, showing a less than absolute correlation between
Wwox expression level and tamoxifen sensitivity, no doubt reflect the heterogeneity of expression of most proteins in individual breast cancers. In addition, ErbB2 expression in the same panel of breast cancers was highly related to Wwox levels (p=0.03); specifically, all the cancers (25/25) showing ErbB2 overexpression (moderate to strong membranous expression in >10% of cells) had very low Wwox levels; 18/25 (72%)
ErbB2 overexpressing cases were tamoxifen resistant, confirming that low Wwox expression correlates with high ErbB2 expression and tamoxifen resistance (p=0.002).
Thus, determination of Wwox expression level, in addition to ER and ErbB2, in breast cancers can help to predict which cases will not respond to tamoxifen treatment.
104 Figure 5.10 Quantitative relationship between Wwox expression level and tamoxifen sensitivity. A: MCF7 cells exhibit the highest level of Wwox expression and are very sensitive to tamoxifen treatment. T47D and BT-474 breast cancer cells, although ERα+, express reduced Wwox levels and show partial resistance to tamoxifen. Wwox expression levels from Western Blots were quantified by densitometry analysis and listed below the graph. B: Growth of tamoxifen treated cells (1 µM) after WWOX knock down by different concentrations (10, 20, 40, 50 nM) of WWOX siRNA oligos. Scrambled
WWOX siRNA (50 nM) served as control. Wwox knock down by >80% is required for complete tamoxifen resistance. C: 63% Wwox knock down allows activation of the PKA pathway, as determined by protein kinase A activity assay; ~85% knock down leads to activation of the ErbB2 pathway, determined by pTyr1248-ErbB2 levels. D: Wwox
(R2=0.9491) and PKA-RIα (R2=0.8728) levels are directly related to tamoxifen sensitivity, while ErbB2 (R2=-0.7902) level is inversely related to tamoxifen sensitivity, as determined by the summary of results of the various assays of cellular activity and expression levels (R value, the correlation coefficient). E: Fulvestrant effectiveness (5 days treatment with 1 nM) is not affected by modulation of Wwox levels by RNAi methods in MCF7 cells. 1. Control; 2. Scrambled siRNA (50 nM); 3. WWOX siRNA
(50 nM); 4. scrambled shRNA plasmid; 5. WWOX shRNA plasmid; 6. MCF7 NSA (non silencing) stable clone; 7. MCF7 W2A (WWOX shRNA) stable clone.
105
106
Figure 5.11 WWOX shRNA stable clones. A) Wwox is down-regulated to various levels in 4 different stable MCF7 clones. B) Clone W2A cells, with 70% Wwox down- regulation, have become significantly resistant (p < 0.05). C) With 70% Wwox down- regulation both PKA and ErbB2 pathways are activated.
107
Wwox expression and tamoxifen response in patients (p = 0.01)
Wwox expression Low Moderate High
13 (24.5%) 22 (41.5%) 18 (34%) Tamoxifen sensitive (equals 41% of all (equals 54% of all (equals 82% of all low expressors) moderate expressors) high expressors)
19 (45.25%) 19 (45.25%) 4 (9.5%) Tamoxifen resistant (equals 59% of all (equals 46% of all (equals 18% of all low expressors) moderate expressors) high expressors)
Figure 5.12 Correlation of Wwox protein levels and tamoxifen sensitivity in breast cancer tissues. A: Strong cytoplamic Wwox expression in a tamoxifen sensitive case (x400); B and C reduced Wwox expression in tamoxifen resistant tumors (x400); D: highly reduced Wwox expression in a tamoxifen resistant case (x400). (e) In the entire study group (N=95), 32 (33.7%) cases exhibited highly reduced, 41 cases (43.2%) reduced and 22 cases (23.2%) strong Wwox expression. Comparison of Wwox expression score with tamoxifen response showed significantly more frequent reduced Wwox levels in tamoxifen resistant cases (p=0.01).
108 5.5 Discussion
We have shown that Wwox down-regulation abrogates tamoxifen sensitivity in breast cancer cells in vitro and in vivo. Three different RNAi approaches were used to down-regulate Wwox expression in breast cancer cells and results consistently showed that Wwox expression levels were highly correlated with tamoxifen sensitivity. The
MCF7 cell line is a standard in vitro model for hormone-sensitive breast cancer and has been widely used to study antiestrogen responsiveness (134-135). Consequently, we chose this cell line to investigate the association of Wwox expression with acquired resistance to tamoxifen, and found that Wwox was highly down-regulated in MCF7 cells selected for resistance to tamoxifen. We concluded that loss of Wwox expression initiates
ErbB2 and PKA tamoxifen resistance pathways. Based on genome-wide gene expression analysis, one major change in MCF7-OHTr cells is the altered expression of estrogen- responsive genes. For example, subsets of estrogen activated or suppressed genes become constitutively activated or suppressed. These alterations in estrogen responsive genes could be caused by aberrant interplay between ERα and growth factor signaling pathways (136). As illustrated in the model shown in Figure 5.13, we propose that cytoplasmic Wwox binds the Ap2γ and Ap2α transcription factors, preventing them from entering the nucleus to activate or repress specific gene targets; when Wwox is lost, the cell loses two pathways implicated in tamoxifen sensitivity. In a variety of in vitro and in vivo breast cancer models, ErbB2 overexpression correlates with tamoxifen resistance
(118-119); the Ap2γ transcription factor is a known activator of ERBB2 transcription and there are several binding sites for Ap2γ in the ERBB2 promoter region (131). We have shown that in tamoxifen sensitive cells, Wwox binds Ap2γ, inhibiting translocation to the 109 nucleus and transcriptional activation of ERBB2, consistent with a previous study suggesting Ap2γ nuclear localization in ErbB2 positive tumors (137) (Figure 5.13A).
Conversely, in tamoxifen resistant cells, Wwox expression is highly reduced, releasing
Ap2γ to translocate to the nucleus and activate the ERBB2 gene (Figure 5.13B).
Activation of ErbB2 was tested by luciferase assay and phosphorylation of tyrosine 1248, which is related to activation of ErbB2 and downstream MAPK pathways contributing to tamoxifen resistance (119,138). PKA can be activated by down-regulation of PKA-RIα, an inhibitory subunit of holo-PKA. Wwox also binds the Ap2α transcription factor in the cytoplasm, preventing transcriptional activity. We have shown that Ap2α is a transcriptional repressor of PKA-RIα. According to this scenario, Wwox binds Ap2α and
PKA-RIα is expressed; PKA-RIα inhibits ERα phosphorylation by PKA and the cells are tamoxifen sensitive. Conversely, loss of Wwox releases Ap2α to translocate to the nucleus to repress PKA-RIα expression. Repression of the inhibitory PKA-RIα allows
PKA catalytic subunits to phosphorylate ERα and induce tamoxifen resistance.
Thus, Wwox, through interaction with Ap2γ and Ap2α, regulates ErbB2 and PKA signaling pathways. PKA has been shown to abrogate tamoxifen effectiveness by phosphorylating ERα at serine 305 (123). A recent report (139) has shown that phosphorylation of ERα serine-305 triggers phosphorylation of serine 118 residue, contributing toward loss of the anti-estrogenic effect of tamoxifen.
110 Figure 5.13 Model of Wwox-mediated tamoxifen resistance. Illustrations of the proposal that loss of Wwox protein initiates the tamoxifen resistance pathway by regulating PKA and ErbB2 signaling pathways. A: Wwox, a cytoplasmic protein, binds the Ap2γ and Ap2α transcription factors and prevents them from entering the nucleus to activate or repress expression of specific gene targets; B: when Wwox is lost, the cell loses a critical modulator of signals affecting cell growth, including two pathways implicated in tamoxifen resistance. i) Ap2γ transcription factor is a known activator of
ErbB2 transcription and in tamoxifen sensitive cells Wwox binds Ap2γ, inhibiting translocation to the nucleus and transcriptional activation of ErbB2; in tamoxifen resistant cells Wwox is down-modulated, releasing Ap2γ to translocate to the nucleus and activate the ERBB2 gene. ii) PKA can be activated by down-regulation of PKA-RIα, an inhibitory subunit of holo-PKA. Ap2α is a transcriptional repressor of PKA-RIα.
According to the model, loss of Wwox in the cytoplasm releases Ap2α to translocate to the nucleus and repress PKA-RIα expression. Repression of the inhibitory PKA-RIα allows the PKA catalytic subunits to phosphorylate ERα and induce tamoxifen resistance.
111
112 Our findings show that Wwox loss is an early step in development of tamoxifen resistance, resulting in dysregulation of several signal pathways that cross-talk with ERα.
The quantitative relationship between Wwox level and tamoxifen response suggested that Wwox expression must be highly reduced (70-80%) to affect its downstream pathways. Wwox levels were also significantly reduced in tamoxifen resistant breast cancers compared to tamoxifen sensitive cancers (p<0.01). High Wwox expression levels in breast cancers can predict tamoxifen sensitivity and, as in our in vitro experiments, Wwox must be reduced by >75% to result in tamoxifen resistance in breast cancers in vivo, perhaps partially due to the low binding affinity of Wwox for
Ap2α, which allows some repression of PKA-RIα when the Wwox expression level is moderate to low. Because these observations are based largely on in vitro data, it is essential to examine the relationship of these proteins in larger panels of tamoxifen sensitive and resistant breast cancers.
Wwox is frequently down-regulated in breast cancer cells by DNA hypermethylation in the promoter region (74), so use of methyltransferase inhibitors could allow re-expression of Wwox, as well as other gene products, and restoration of tamoxifen sensitivity (140). This has important therapeutic implications, suggesting for the first time that combining low doses of demethylating agents with tamoxifen treatment may prevent acquired tamoxifen resistance. It is already known, in cases of de novo tamoxifen resistance, that epigenetic reactivation of ERα can sensitize breast cancer cells to tamoxifen (141-142). Wwox levels do not affect fulvestrant sensitivity of breast cancers and most tamoxifen resistant cancers respond to this drug (123). Thus, determination of Wwox levels in breast cancers may prove to aid in prediction of 113 tamoxifen resistance in breast cancers and may identify breast cancers for which fulvestrant treatment would be more effective. This is important considering the varying results of microarray analyses aimed at predicting a tamoxifen resistance signature (143-
145).
In summary, we have found that loss of Wwox expression initiates tamoxifen resistance pathways, and describe novel mechanisms connecting Wwox loss with ErbB2 and PKA pathways to tamoxifen resistance. The proposed mechanisms illustrate how
ErbB2 is activated and PKA-RIα repressed in tamoxifen resistant breast cancers. ErbB2 and PKA pathways are important in breast cancer, not only because of the connection with tamoxifen resistance (146-149). More than 25% of breast tumors overexpress ErbB2 and it was recently proposed that ErbB2 overexpression is related to resistance to aromatase inhibitors (150). Also, illustration of the novel mechanisms of ErbB2 and PKA regulation provides targets for possible therapeutic intervention. Results of the immunohistochemical analyses of Wwox expression levels in breast cancers with known tamoxifen response has supported the proposal that level of Wwox expression is correlated with tamoxifen response and provides the rationale for a detailed assessment of subcellular location and expression level of Wwox and Wwox interactors in large panels of tamoxifen sensitive and refractory breast cancers.
Publication: Iliopoulos D, Guler G, Guler N, Druck T, Fan M, Han SY, Fabbri M, Semba S, Di Leva G, Weigel R, Nephew K and Huebner K. Wwox is a Critical Mediator of Tamoxifen Response (submitted).
114
CHAPTER 6
CONCLUSIONS AND FUTURE DIRECTIONS
The results of studies described in chapters 2-5 have strongly confirmed the tumor suppressor function of the WWOX gene product, especially in breast and lung cancer and have illustrated mechanisms of modulation of Wwox expression as well as pathways through which Wwox protein functions in cancer cells.
Specifically, our studies have shown that: 1) the WWOX tumor suppressor gene is frequently down-regulated or lost by LOH and/or regulatory region DNA methylation in breast and lung cancer. The WWOX promoter was heavily methylated in cancers which was related to its expression. We suggest that WWOX methylation specific amplification will be able to distinguish cancer DNA from DNA of adjacent and normal tissue enriching a panel of DNA methylation markers. 2) In addition we have found that restoration of Wwox expression was able to induce apoptosis in breast and lung cancer cells without affecting Wwox positive cells. 3) Also Wwox restoration was highly effective in gene therapy of cancer xenografts transduced by adenoviral vector or through a drug-inducible system. 4) Furthermore, due to the epigenetic regulation of Wwox, restoration of its expression by demethylating agents and histone deacetylase inhibitors was able to suppress tumor growth in vitro and in vivo. According to preclinical studies,
115 we suggest the therapeutic potential of restoration of tumor suppressor genes through epigenetic modulation. After studying Wwox regulatory mechanisms and its therapeutic potential we were interested to study in more detail its function in breast cancer cells.
5) We have found that Wwox regulation of ErbB2 and PKA signaling pathways through
Ap2 transcription factors mediates tamoxifen effectiveness, a drug that has been used for more than 20 years in the clinic. Loss of Wwox initiates tamoxifen resistance through release of Ap2 transcription factors to the nucleus where Ap2α down-regulates PKA-
RIα and Ap2γ up-regulates ErbB2 expression. Up-regulation of Wwox expression using an expression vector or demethylating agents restored tamoxifen sensitivity. Our Wwox- mediated tamoxifen resistance model was validated by the examination of breast cancer samples where Wwox expression was significantly reduced in tamoxifen resistant tumors. These results will provide new targets for therapeutic intervention in drug resistant breast cancers. Some implications of these conclusions are described in detail in the following section.
6.1 WWOX methylation as a potential biomarker for lung and breast cancer
The renaissance of methylation studies has generated considerable interest for cancer researchers because: 1) methylation of CpG islands in gene regulatory regions, in combination with chromatin remodelling, is involved in downregulation of expression of genes, including tumor suppressor genes; 2) methylation of specific CpG islands is easily detected in tissues and body fluids of individuals with cancer (61,151-153), or at high risk for cancer development, so that specific gene methylation patterns can be useful in
116 diagnosis or prognosis; 3) epigenetic and chromatin remodelling marks can be reversed by specific agents or inhibitors, suggesting such inhibitors as therapeutic agents (79,
154).
We have investigated the methylation status and expression levels of WWOX and
FHIT in breast, lung and bladder cancer. Both genes seem to be coordinately inactivated in cancers, frequently due to DNA hypermethylation in regulatory regions (74). Many studies have shown that, although the abnormal epigenetic silencing of genes can occur at any time during tumor progression, it occurs most frequently during the early stages of the neoplastic process, such as the pre-cancerous stages of tumour development (3,83-
85,94-98,106). The differential patterns of WWOX and FHIT methylation that we have observed in neoplastic vs. adjacent non neoplastic tissues, suggests that targeted MSP amplification could be useful in following treatment or prevention protocols. Methylation of regulatory regions of many genes has been reported in cancer cells, but which of these methylation marks will be most useful in diagnostic or prognostic clinical trials, or as surrogate markers in preclinical/clinical prevention and therapy trials of specific cancers, is only beginning to be defined.
6.2 WWOX gene therapy in lung and breast cancer
The majority of cancer patients present with advanced disease. Often, the primary tumor mass can be resected, although surgery usually cannot be used to remove disseminated metastases that are frequently responsible for ultimate patient demise.
Agents capable of treating such disseminated disease must have a high therapeutic index,
117 showing selective toxicity against cancer tissue while sparing normal cells. Gene therapy is one promising approach because targeting to cancer cells can be combined with intracellular mechanisms regulating selective expression to achieve good tumor specificity (156). Tumor-selective activity can be used to regulate toxicity through the amplification of a gene product from even a single gene copy (155-156).
Adenovirus is a promising vector for cancer gene therapy, and the development of a form that can be administered systemically and targeted to disseminated tumors is a major challenge facing cancer gene therapy (156-157). However, there are several issues to be addressed before this is likely to be a feasible approach. Most obvious is the problem of rapid clearance of virus from the blood, widely reported and discussed by several authors who have used a range of techniques to prevent unwanted entry of the virus into nontarget tissues (158-159). Second is the question of target selectivity, where effective expression in target tissue must be combined with avoiding unwanted infection of healthy tissues in order to prevent unnecessary immune provocation and vector depletion (160).
We have shown that Wwox and Fhit protein expression is lost or reduced in early stages of lung and breast carcinogenesis (28,74). Restoration of these tumor suppressors by adenoviral vectors suppressed cell growth and induced apoptosis in vitro and in vivo.
Recombinant adenovirus injection in lung cancer phase I trials (71) has demonstrated safety and feasibility and phase I/II clinical trials are currently recruiting patients to evaluate toxicity and efficacy of gene therapy with recombinant adenoviruses (72). There are several studies published showing the efficiency of Fhit restoration in suppressing 118 tumor growth in animal models (87,89-92,102). It is important to conduct more studies in order to determine if also Wwox restoration can suppress tumorigenicity in vivo especially in accessible cancers.
In addition, we have found that Wwox down-regulation in several cancers is due to DNA hypermethylation (74). Currently, there are several demethylating agents used in preclinical and clinical trials (105-110). Demethylation and reactivation of several tumor suppressor genes, including WWOX, resulted in inhibition of tumor growth in vitro and in vivo (82,161). According to our studies, even very low doses of demethylating agents can block tumor growth. The specific design of optimal protocols (doses, schedule) using these agents will minimize the side effects and maximize specificity. WWOX gene transfer in the future may be more feasible after development of nanotechnologies that can target tumor cells more efficiently and specifically (162-163).
6.3 Wwox levels can predict tamoxifen sensitivity
Although many patients benefit from tamoxifen treatment, half of breast tumors that recur after therapy are resistant to tamoxifen. Understanding mechanisms of tamoxifen resistance might identify protein markers for nonresponsive cancers, as well as tumors that are acquiring resistance, before emergence of more aggressive cancer cells.
We have found that: 1) the Wwox tumor suppressor can mediate the tamoxifen response through regulation of protein kinase A and ErbB2 signaling pathways, and 2) high Wwox expression level predicted tamoxifen sensitivity in a cohort of breast cancer cases (p<0.01). The results also imply that epigenetic reactivation of Wwox could
119 sensitize tamoxifen resistant cells, suggesting possible epigenetic and hormonal combination therapy in patients with acquired tamoxifen resistance.
As we described previously, Wwox is frequently down-regulated in breast cancer cells by DNA hypermethylation in the promoter region (74), so use of methyltransferase inhibitors could allow re-expression of Wwox, as well as other gene products, and restoration of tamoxifen sensitivity (140). This has important therapeutic implications, suggesting for the first time that combining low doses of demethylating agents with tamoxifen treatment may prevent acquired tamoxifen resistance. Thus, determination of
Wwox levels in breast cancers may prove to aid in prediction of tamoxifen resistance in breast cancers and may identify breast cancers for which fulvestrant treatment would be more effective. This is important considering the varying results of microarray analyses aimed at predicting a tamoxifen resistance signature (143-145).
6.4 Wwox as a regulator of major signaling pathways in breast cancer
According to our results and results of other groups, Wwox appears to be a critical mediator of cell signaling pathways in breast cancer cells. Wwox harbors two
WW domains, which are known to mediate protein-protein interactions, and an SDR domain, which may play an important role in hormone-related cancers such as breast and prostate cancer. It has already been shown that Wwox interacts with several proteins in breast cancer cells, supporting the idea of its role as a signal mediator. We believe that
Wwox is a negative regulator of proteins that increase cell proliferation, consistent with its role as a tumor suppressor gene. Wwox down-regulation in several cancers allows its
120 interactors to activate their downstream targets, which increase cell proliferation and tumorigenesis. Probably in different cell types or cancers Wwox, interacts with different proteins. It is important to perform proteomic analyses in order to identify and characterize Wwox-containing protein complexes in order to delineate breast cancer signaling pathways.
121
BIBLIOGRAPHY
1. Huebner K, Croce CM. FRA3B and other common fragile sites: the weakest links. Nat Rev Cancer. 1: 214-21 (2001).
2. El Achkar E, Gerbault-Seureau M, Muleris M, Dutrillaux B, Debatisse M. Premature condensation induces breaks at the interface of early and late replicating chromosome bands bearing common fragile sites. Proc Natl Acad Sci U S A. 102: 18069-74 (2005).
3. Yunis JJ, Soreng AL. Constitutive fragile sites and cancer. Science. 226: 1199-204 (1984).
4. Casper AM, Nghiem P, Arlt MF and Glover TW. ATR regulates fragile site stability. Cell. 111: 779–789 (2002).
5. Mishmar D, Rahat A, Scherer SW, Nyakatura G, Hinzmann B and Kohwi Y et al., Molecular characterization of a common fragile site (FRA7H) on human chromosome 7 by the cloning of a simian virus 40 integration site. Proc. Natl Acad. Sci. USA. 95: 8141– 8146 (1998).
6. Zlotorynski E, Rahat A, Skaug J, Ben-Porat N, Ozeri E and Hershberg R. et al., Molecular basis for expression of common and rare fragile sites. Mol. Cell Biol. 23: 7143-51 (2003).
7. Arlt MF, Xu B, Durkin SG, Casper AM, Kastan MB and Glover TW. BRCA1 is required for common-fragile-site stability via its G2/M checkpoint function. Mol. Cell Biol. 24: 6701-9 (2004).
122 8. Howlett NG, Taniguchi T, Durkin SG, D'Andrea AD and Glover TW. The Fanconi anemia pathway is required for the DNA replication stress response and for the regulation of common fragile site stability. Hum. Mol. Genet. 14: 693-701 (2005).
9. Musio A, Montagna C, Mariani T, Tilenni M, Focarelli ML and Brait L, Indino E, Benedetti PA, Chessa L, Albertini A, Ried T, Vezzoni P. SMC1 involvement in fragile site expression. Hum. Mol. Genet. 14: 525-533 (2005).
10. Wang L, Darling J, Zhang JS, Huang H, Liu W, Smith DI. Allele-specific late replication and fragility of the most active common fragile site, FRA3B. Hum. Mol. Genet. 8: 431-7 (1999).
11. Palakodeti A, Han Y, Jiang Y, Le Beau MM. The role of late/slow replication of the FRA16D in common fragile site induction. Genes Chromosomes Cancer 39: 71-6 (2004).
12. Hellman A, Rahat A, Scherer SW, Darvasi A, Tsui LC, Kerem B. Replication delay along FRA7H, a common fragile site on human chromosome 7, leads to chromosomal instability. Mol. Cell Biol. 20: 4420-7 (2000).
13. Boldog FL, Waggoner B, Glover TW, Chumakov I, Le Paslier D, Cohen D, Gemmill RM, Drabkin HA. Integrated YAC contig containing the 3p14.2 hereditary renal carcinoma 3;8 translocation breakpoint and the fragile site FRA3B. Genes Chromosomes Cancer. 11: 216-21 (1994).
14. Wilke CM, Guo SW, Hall BK, Boldog F, Gemmill RM, Chandrasekharappa SC, Barcroft CL, Drabkin HA, Glover TW. 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: 319-26 (1994).
15. Paradee W, Wilke CM, Wang L, Shridhar R, Mullins CM, Hoge A, Glover TW, Smith DI. 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 :87-93 (1996).
16. Cohen AJ, Li FP, Berg S, Marchetto DJ, Tsai S, Jacobs SC, Brown RS. Hereditary renal-cell carcinoma associated with a chromosomal translocation. N Engl J Med. 301: 592-5 (1979).
123 17. Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T, Croce CM, Huebner K. 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: 587-97 (1996).
18. Zimonjic DB, Druck T, Ohta M, Kastury K, Croce CM, Popescu NC, Huebner K. Positions of chromosome 3p14.2 fragile sites (FRA3B) within the FHIT gene. Cancer Res. 57: 1166-70 (1997).
19. Sozzi G, Tornielli S, Tagliabue E, Sard L, Pezzella F, Pastorino U, Minoletti F, Pilotti S, Ratcliffe C, Veronese ML, Goldstraw P, Huebner K, Croce CM, Pierotti MA. Absence of Fhit protein in primary lung tumors and cell lines with FHIT gene abnormalities. Cancer Res. 57: 5207-12 (1997).
20. Huebner K, Croce CM. Cancer and the FRA3B/FHIT fragile locus: it's a HIT. Br J Cancer. 10: 1501-6 (2003).
21. Thiagalingam S, Lisitsyn NA, Hamaguchi M, Wigler MH, Willson JK, Markowitz SD, Leach FS, Kinzler KW, Vogelstein B. Evaluation of the FHIT gene in colorectal cancers, Cancer Res. 56: 2936-9 (1996).
22. Le Beau MM, Drabkin H, Glover TW, Gemmill R, Rassool FV, McKeithan TW, Smith DI. An FHIT tumor suppressor gene? Genes Chromosomes Cancer. 21: 281-9 (1998).
23. Bednarek AK, Laflin KJ, Daniel RL, Liao Q, Hawkins KA, Aldaz CM. WWOX, a novel WW domain-containing protein mapping to human chromosome 16q23.3-24.1, a region frequently affected in breast cancer. Cancer Res. 60: 2140-5 (2000).
24. Ludes-Meyers JH, Bednarek AK, Popescu NC, Bedford M, Aldaz CM. WWOX, the common chromosomal fragile site, FRA16D, cancer gene. Cytogenet Genome Res. 100 101-10 (2003).
25. Bednarek AK, Keck-Waggoner CL, Daniel RL, Laflin KJ, Bergsagel PL, Kiguchi K, Brenner AJ, Aldaz CM. WWOX, the FRA16D gene, behaves as a suppressor of tumor growth. Cancer Res. 61: 8068-73 (2001).
124 26. Paige AJ, Taylor KJ, Stewart A, Sgouros JG, Gabra H, Sellar GC, Smyth JF, Porteous DJ, Watson JE. A 700-kb physical map of a region of 16q23.2 homozygously deleted in multiple cancers and spanning the common fragile site FRA16D. Cancer Res 60: 1690-7 (2000).
27. Mangelsdorf M, Ried K, Woollatt E, Dayan S, Eyre H, Finnis M, Hobson L, Nancarrow J, Venter D, Baker E, Richards RI. Chromosomal fragile site FRA16D and DNA instability in cancer. Cancer Res. 60: 1683-9 (2000).
28. Guler G, Uner A, Guler N, Han SY, Iliopoulos D, Hauck WW, McCue P, Huebner K. The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinoma. Cancer. 8: 1605-14 (2004).
29. Nunez MI, Ludes-Meyers J, Abba MC, Kil H, Abbey NW, Page RE, Sahin A, Klein- Szanto AJP, Aldaz CM. Frequent loss of WWOX expression in breast cancer: correlation with estrogen receptor status. Breast Cancer Res Treat. 89: 99-105 (2005).
30. G. Guler, A. Uner, N. Güler, S.Y. Han, D. Iliopoulos, P. McCue, K. Huebner, Concordant Loss of Fragile Gene Expression Early in Breast Cancer Development. Pathology International. 55: 471-8 (2005).
31. Hsieh CL, Jones PA. Meddling with methylation. Nat Cell Biol. 5: 502-4 (2003).
32. Yang AS, Doshi KD, Choi SW, Mason JB, Mannari RK, Gharybian V, Luna R, Rashid A, Shen L, Estecio MR, Kantarjian HM, Garcia-Manero G, Issa JP. DNA Methylation Changes after 5-Aza-2'-Deoxycytidine Therapy in Patients with Leukemia. Cancer Res. 66: 5495-5503 (2006).
33. Momparler RL. Epigenetic therapy of cancer with 5-aza-2'-deoxycytidine (decitabine). Semin Oncol. 32 :443-51 (2005).
34. Coradini D, Pellizzaro C, Scarlata I, Zorzet S, Garrovo C, Abolafio G, Speranza A, Fedeli M, Cantoni S, Sava G, Daidone MG, Perbellini A. A novel retinoic/butyric hyaluronan ester for the treatment of acute promyelocytic leukemia: preliminary preclinical results. Leukemia. 20: 785-92 (2006).
125 35. Ryan QC, Headlee D, Acharya M, Sparreboom A, Trepel JB, Ye J, Figg WD, Hwang K, Chung EJ, Murgo A, Melillo G, Elsayed Y, Monga M, Kalnitskiy M, Zwiebel J, Sausville EA. Phase I and pharmacokinetic study of MS-275, a histone deacetylase inhibitor, in patients with advanced and refractory solid tumors or lymphoma. J Clin Oncol. 23: 3912-22 (2005).
36. Ishii H, Mimor K, Inageta T, Murakumo Y, Vecchione A, Mori M, Furukawa Y. Components of DNA damage checkpoint pathway regulate UV exposure-dependent alterations of gene expression of FHIT and WWOX at chromosome fragile sites. Mol Cancer Res. 3:130-8 (2005).
37. Kuroki T, Yendamuri S, Trapasso F, Matsuyama A, Aqeilan RI, Alder H, Rattan S, Cesari R, Nolli ML, Williams NN, Mori M, Kanematsu T, Croce CM. The tumor suppressor gene WWOX at FRA16D is involved in pancreatic carcinogenesis. Clin Cancer Res. 10: 2459-65 (2004).
38. Macias MJ, Wiesner S, Sudol M. WW and SH3 domains, two different scaffolds to recognize proline-rich ligands. FEBS Lett. 513: 30-37 (2002).
39. Ilsley JL, Sudol M, Winder SJ. The WW domain: linking cell signalling to the membrane cytoskeleton. Cell signaling 14: 183-9 (2002).
40. Duax WL, Ghosh D, Pletnev V. Steroid dehydrogenase structures, mechanism of action, and disease. Vitam Horm. 58: 121-48 (2000).
41. Chang NS, Schultz L, Hsu LJ, Lewis J, Su M, Sze CI. 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: 714-23 (2005).
42. Aqeilan RI, Pekarsky Y, Herrero JJ, Palamarchuk A, Letofsky J, Druck T, Trapasso F, Han SY, Melino G, Huebner K, Croce CM. Functional association between Wwox tumor suppressor protein and p73, a p53 homolog. Proc Natl Acad Sci U S A. 101: 4401- 6 (2004).
43. Aqeilan RI, Palamarchuk A, Weigel RJ, Herrero JJ, Pekarsky Y, Croce CM. Physical and functional interactions between the Wwox tumor suppressor protein and the AP- 2gamma transcription factor, Cancer Res. 64: 8256-6 (2004).
126
44. Aqeilan RI, Donati V, Palamarchuk A, Trapasso F, Kaou M, Pekarsky Y, Sudol M, Croce CM. WW domain-containing proteins, WWOX and YAP, compete for interaction with ErbB-4 and modulate its transcriptional function. Cancer Res. 65:6764-72 (2005).
45. Paige AJ, Taylor KJ, Taylor C, Hillier SG, Farrington S, Scott D, Porteous DJ, Smyth JF, Gabra H, Watson JE. WWOX: a candidate tumor suppressor gene involved in multiple tumor types. Proc Natl Acad Sci USA. 98: 11417-22 (2001).
46. Driouch K, Prydz H, Monese R, Johansen H, Lidereau R, Frengen E. Alternative transcripts of the candidate tumor suppressor gene, WWOX, are expressed at high levels in human breast tumors. Oncogene. 21: 1832-40 (2002).
47. Kuroki T, Trapasso F, Shiraishi T, Alder H, Mimori K, Mori M, Croce CM. Genetic alterations of the tumor suppressor gene WWOX in esophageal squamous cell carcinoma. Cancer Res. 62: 2258-60 (2002).
48. Yendamuri S, Kuroki T, Trapasso F, Henry AC, Dumon KR, Huebner K, Williams NN, Kaiser LR, Croce CM. WW domain containing oxidoreductase gene expression is altered in non-small cell lung cancer. Cancer Res. 63: 878-81 (2003).
49. Ingvarsson S, Agnarsson BA, Sigbjornsdottir BI, Kononen J, Kallioniemi OP, Barkardottir RB, Kovatich AJ, Schwarting R, Hauck WW, Huebner K, McCue PA. Reduced Fhit expression in sporadic and BRCA2-linked breast carcinomas. Cancer Res. 59: 2682-9 (1999).
50. Baffa R, Gomella LG, Vecchione A, Bassi P, Mimori K, Sedor J, Calviello CM, Gardiman M, Minimo C, Strup SE, McCue PA, Kovatich AJ, Pagano F, Huebner K, Croce CM. Loss of FHIT expression in transitional cell carcinoma of the urinary bladder. Am J Pathol. 156: 419-24 (2000).
51. Gatalica Z, Lele SM, Rampy BA, Norris BA. The expression of Fhit protein is related inversely to disease progression in patients with breast carcinoma. Cancer. 88: 1378-83 (2000).
52. Maruyama R, Toyooka S, Toyooka KO, Harada K, Virmani AK, Zochbauer-Muller S, Farinas AJ, Vakar-Lopez F, Minna JD, Sagalowsky A, Czerniak B, Gazdar AF.
127 Aberrant promoter methylation profile of bladder cancer and its relationship to clinicopathological features. Cancer Res. 61: 8659-63 (2001).
53. Zochbauer-Muller S, Fong KM, Maitra A, Lam S, Geradts J, Ashfaq R, Virmani AK, Milchgrub S, Gazdar AF, Minna JD. 5' CpG island methylation of the FHIT gene is correlated with loss of gene expression in lung and breast cancer. Cancer Res. 61: 3581- 5 (2001).
54. Kim H, Kwon YM, Kim JS, Lee H, Park JH, Shim YM, Han J, Park J, Kim DH. Tumor-specific methylation in bronchial lavage for the early detection of non-small-cell lung cancer. J Clin Oncol. 22: 2363-70 (2004).
55. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci U S A. 93: 9821-6 (1996).
56. Skopelitou AS, Gloustianou G, Bai M, Huebner K. FHIT gene expression in human urinary bladder transitional cell carcinomas. In Vivo. 15: 169-73 (2001).
57. Tanaka H, Shimada Y, Harada H, Shinoda M, Hatooka S, Imamura M, Ishizaki K. Methylation of the 5' CpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas. Cancer Res. 58: 3429-34 (1998).
58. Esteller M, Sanchez-Cespedes M, Rosell R, Sidransky D, Baylin SB, Herman JG. Detection of aberrant promoter hypermethylation of tumor suppressor genes in serum DNA from non-small cell lung cancer patients. Cancer Res. 59: 67-70 (1999).
59. Yang Q, Nakamura M, Nakamura Y, Yoshimura G, Suzuma T, Umemura T, Shimizu Y, Mori I, Sakurai T, Kakudo K. Two-hit inactivation of FHIT by loss of heterozygosity and hypermethylation in breast cancer. Clin Cancer Res. 8: 2890-3 (2002).
60. Huebner K, Garrison PN, Barnes LD, Croce CM. The role of the FHIT/FRA3B locus in cancer. Ann Rev Genet. 32: 7-31 (1998).
61. Palmisano WA, Divine KK, Saccomanno G, Gilliland FD, Baylin SB, Herman JG, Belinsky SA. Predicting lung cancer by detecting aberrant promoter methylation in sputum. Cancer Res. 60: 5954-8 (2000). 128
62. Soria JC, Rodriguez M, Liu DD, Lee JJ, Hong WK, Mao L. Aberrant promoter methylation of multiple genes in bronchial brush samples from former cigarette smokers. Cancer Res. 62: 351-5 (2000).
63. Muller HM, Widschwendter A, Fiegl H, Ivarsson L, Goebel G, Perkmann E, Marth C, Widschwendter M. DNA methylation in serum of breast cancer patients: an independent prognostic marker. Cancer Res. 63: 7641-5 (2003).
64. Tang X, Khuri FR, Lee JJ, Kemp BL, Liu D, Hong WK, Mao L. Hypermethylation of the death-associated protein (DAP) kinase promoter and aggressiveness in stage I non- small-cell lung cancer. J Natl Cancer Inst. 92: 1511-6 (2000).
65. Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer Statistics. 51: 15-36 (2001).
66. Roth JA, Nguyen D, Lawrence DD, Kemp BL, Carrasco CH, Ferson DZ, Hong WK, Komaki R, Lee JJ, Nesbitt JC, Pisters KM, Putnam JB, Schea R, Shin DM, Walsh GL, Dolormente MM, Han CI, Martin FD, Yen N, Xu K, Stephens LC, McDonnell TJ, Mukhopadhyay T, Cai D. Retrovirus-mediated wild-type p53 gene transfer to tumors of patients with lung cancer. Nat. Med. 2: 985-991 (1996).
67. Nemunaitis J, Swisher SG, Timmons T, Connors D, Mack M, Doerksen L, Weill D, Wait J, Lawrence DD, Kemp BL, Fossella F, Glisson BS, Hong WK, Khuri FR, Kurie JM, Lee JJ, Lee JS, Nguyen JC, Nesbitt JC, Perez-Soler R, Pisters KM, Putnam JB, Richli WR, Shin DM, Walsh GL, Merrit J, Roth J. Adenovirus-mediated p53 gene transfer in sequence with cisplatin to tumors of patients with non-small-cell lung cancer. J. Clin. Oncol. 3: 609-622 (2000).
68. Roth JA, Swisher SG, Merritt JA, Lawrence DD, Kemp BL, Carrasco CH, El-Naggar AK, Fossella FV, Glisson BS, Hong WK, Khurl FR, Kurie JM, Nesbitt JC, Pisters K, Putnam JB, Schrump DS, Shin DM, Walsh GL. Gene therapy for non-small cell lung cancer: a preliminary report of a phase I trial of adenoviral p53 gene replacement. Semin. Oncol. 25: 33-37 (1998).
69. Weill D, Mack M, Roth J, Swisher S, Proksch S, Merritt J, Nemunaitis, J. Adenoviral-mediated p53 gene transfer to non-small cell lung cancer through endobronchial injection. Chest 118: 966-970 (2000).
129 70. Swisher SG, Roth JA, Nemunaitis J, Lawrence DD, Kemp BL, Carrasco CH, Connors DG, El-Naggar AK, Fossella F, Glisson BS, Hong WK, Khuri FR, Kurie JM, Lee JJ, Lee JS, Mack M, Merrit JA, Nguyen DM, Nesbitt JC, Perez-Soler R, Pisters KM, Putnam JB Jr, Richli WR, Savin M, Schrump DS, Shin DM, Shulkin A, Walsh GL, Wait J, Weill D, Waugh, M.K. Adenovirus-mediated p53 gene transfer in advanced non-small-cell lung cancer. J. Natl. Cancer Inst. 91: 763-771 (1999).
71. Griscelli F, Opolon P, Saulnier P, Mami-Chouaib F, Gautier E, Echchakir H, Angevin E, Le Chevalier T, Bataille V, Squiban P, Tursz T, Escudier B. Recombinant adenovirus shedding after intratumoral gene transfer in lung cancer patients. Gene Ther. 10: 386-395 (2003).
72. Carbone DP, Adak S, Schiller J, Slovis B, Kubba S, Coffee K, Worrell J, Thet L, Krozely P, Johnson D. Proc. Am. Soc. Clin. Oncol. 22: 620 (abstr. 2492) (2003).
73. Sozzi G, Pastorino U, Moiraghi L, Tagliabue E, Pezzella F, Girelli C, Tornelli S, Sard L, Huebner K, Pierotti MA, Croce CM, Pilotti S. Loss of FHIT function in lung cancer and preinvasive bronchial lesions. Cancer Res. 58, 5032-5037 (1998).
74. Iliopoulos D, Guler G, Han SY, Johnston D, Druck T, McCorkell KA, Palazzo J, McCue PA, Baffa R, Huebner K. Fragile genes as biomarkers: epigenetic control of WWOX and FHIT in lung, breast and bladder cancer. Oncogene. 24: 1625-1633 (2005).
75. Roz L, Gramegna M, Ishii H, Croce CM, Sozzi G. Restoration of fragile histidine triad (FHIT) expression induces apoptosis and suppresses tumorigenicity in lung and cervical cancer cell lines. Proc Natl Acad Sci USA. 99: 3615-20 (2002).
76. Ingrosso D, Cimmino A, Perna AF, Masella L, De Santo NG, De Bonis M, Vacca M, D’Esposito M, D’Urso M, Galletti P, Zappia V. Folate treatment and unbalanced methylation and changes of allelic expression induced by hyperhomocysteinaemia in patients with uraemia. Lancet. 361: 1693-1699 (2003).
77. McGregor F, Muntoni A, Fleming J, Brown J, Felix DH, MacDonald DG, Parkinson EK, Harrison PR. Molecular changes associated with oral dysplasia progression and acquisition of immortality: potential for its reversal by 5-azacytidine. Cancer Res. 16: 4757-4766 (2002).
130 78. Hennessy BT, Garcia-Manero G, Kantarjian HM, Giles FJ. DNA methylation in haematological malignancies: the role of decitabine. Expert Opin. Invest. Drugs 12: 1985-1993 (2003).
79. Takai N, Desmond JC, Kumagai T, Gui D, Said JW, Whittaker S, Miyakawa I, Koeffler HP. Histone deacetylase inhibitors have a profound antigrowth activity in endometrial cancer cells. Clin. Cancer Res. 10, 1141-1149 (2004).
80. Lerman MI, Minna JD. The 630-kb lung cancer homozygous deletion region on human chromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressor genes. Cancer Res. 60: 6116-33 (2000).
81. Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG, Inoue H, Tornielli S, Pilotti S, De Gregorio L, Pastorino U, Pierotti MA, Ohta M et al. The FHIT gene 3p14.2 is abnormal in lung cancer. Cell. 85: 17-26 (1996).
82. Fabbri M, Iliopoulos D, Trapasso F, Aqeilan RI, Cimmino A, Zanesi N, Yendamuri S, Han SY, Amadori D, Huebner K, Croce CM. WWOX gene restoration prevents lung cancer growth in vitro and in vivo. Proc Natl Acad Sci U S A. 102: 15611-6 (2005).
83. Damann R, Strunnikova M, Schardarsurengin U, Rastetter M, Papritz M, Hattenhorst, UF, Silber RE, Burdach S, Hansen G. CpG island methylation and expression of tumour- associated genes in lung carcinoma. Eur. J Cancer. 41: 1223-1236 (2005).
84. Pfeifer GP, Damann R. Methylation of the tumor suppressor gene RASSF1A in human tumors. Biochemistry (Mosc) 70: 576-583 (2005).
85. Grote HJ, Schmiemann V, Geddert H, Bocking A, Kappes R, Gabbert HE, Sarbia M. Methylation of RAS association domain family protein 1A as a biomarker of lung cancer. Cancer, epub Jan, 2006.
86. Dumon KR, Ishii H, Vecchione A, Trapasso F, Baldassarre G, Chakrani F, Druck T, Rosato EF, Williams NN, Baffa R, During MJ, Huebner K et al. Fragile histidine triad expression delays tumor development and induces apoptosis in human pancreatic cancer. Cancer Res. 61: 4827-36 (2001).
131 87. Dumon KR, Ishii H, Fong LY, Zanesi N, Fidanza V, Mancini R, Vecchione A, Baffa R, Trapasso F, During MJ, Huebner K, Croce CM. FHIT gene therapy prevents tumor development in Fhit-deficient mice. Proc Natl Acad Sci U S A. 98: 3346-51 (2001).
88. Ji L, Fang B, Yen N, Fong K, Minna JD, Roth JA. Induction of apoptosis and inhibition of tumorigenicity and tumor growth by adenovirus vector-mediated fragile histidine triad (FHIT) gene overexpression. Cancer Res. 59: 3333-9 (1999).
89. Siprashvili Z, Sozzi G, Barnes LD, McCue P, Robinson AK, Eryomin V, Sard L, Tagliabue E, Greco A, Fusetti L, Schwartz G, Pierotti MA et al. Replacement of Fhit in cancer cells suppresses tumorigenicity. Proc Natl Acad Sci U S A 1997;94:13771-6.
90. Sevignani C, Calin GA, Cesari R, Sarti M, Ishii H, Yendamuri S, Vecchione A, Trapasso F, Croce CM. Restoration of fragile histidine triad (FHIT) expression induces apoptosis and suppresses tumorigenicity in breast cancer cell lines. Cancer Res. 63: 1183- 7 (2003).
91. Ishii H, Dumon KR, Vecchione A, Trapasso F, Mimori K, Alder H, Mori M, Sozzi G, Baffa R, Huebner K, Croce CM. Effect of adenoviral transduction of the fragile histidine triad gene into esophageal cancer cells. Cancer Res. 61: 1578-84 (2001).
92. Ishii H, Zanesi N, Vecchione A, Trapasso F, Yendamuri S, Sarti M, Baffa R, During MJ, Huebner K, Fong LY, Croce CM. Regression of upper gastric cancer in mice by FHIT gene delivery. FASEB Journal. 17: 1768-70 (2003).
93. Guler G, Iliopoulos D, Han SY, Fong LY, Lubet RA, Grubbs CJ, Huebner K. Hypermethylation patterns in the Fhit regulatory region are tissue specific. Mol Carcinog 43: 175-81 (2005).
94. Tomizawa Y, Iijima H, Nomoto T, Iwasaki Y, Otani Y, Tsuchiya S, Saito R, Dobashi K, Nakajima T, Mori M. Clinicopathological significance of aberrant methylation of RARbeta2 at 3p24, RASSF1A at 3p21.3, and FHIT at 3p14.2 in patients with non-small cell lung cancer. Lung Cancer 2004;46:305-12.
95. Maruyama R, Sugio K, Yoshino I, Maehara Y, Gazdar AF. Hypermethylation of FHIT as a prognostic marker in nonsmall cell lung carcinoma. Cancer 100: 1472-7 (2004).
132 96. Zheng S, Ma X, Zhang L, Gunn L, Smith MT, Wiemels JL, Leung K, Buffler PA, Wiencke JK. Hypermethylation of the 5' CpG island of the FHIT gene is associated with hyperdiploid and translocation-negative subtypes of pediatric leukemia. Cancer Res 64:2000-6 (2004).
97. Ito M, Ito G, Kondo M, Uchiyama M, Fukui T, Mori S, Yoshioka H, Ueda Y, Shimokata K, Sekido Y. Frequent inactivation of RASSF1A, BLU, and SEMA3B on 3p21.3 by promoter hypermethylation and allele loss in non-small cell lung cancer. Cancer Lett. 225:131-9 (2005).
98. Ulivi P, Zoli W, Calistri D, Fabbri F, Tesei A, Rosetti M, Mengozzi M, Amadori D. p16(INK4A) and CDH13 hypermethylation in tumor and serum of non-small cell lung cancer patients. J Cell Physiol. 206: 611-5 (2006).
99. Ottey M, Han SY, Druck T, Barnoski BL, McCorkell KA, Croce CM, Raventos- Suarez C, Fairchild CR, Wang Y, Huebner K. Fhit-deficient normal and cancer cells are mitomycin C and UVC resistant. Br J Cancer. 91: 1669-77 (2004).
100. Suzuki M, Sunaga N, Shames DS, Toyooka S, Gazdar AF, Minna JD. RNA interference-mediated knockdown of DNA methyltransferase 1 leads to promoter demethylation and gene re-expression in human lung and breast cancer cells. Cancer Res. 64: 3137-43 (2004).
101. Ishii H, Dumon KR, Vecchione A, Fong LY, Baffa R, Huebner K, Croce CM. Potential cancer therapy with the fragile histidine triad gene: review of the preclinical studies. JAMA. 286: 2441-9 (2001).
102. Zanesi N, Mancini R, Sevignani C, Vecchione A, Kaou M, Valtieri M, Calin GA, Pekarsky Y, Gnarra JR, Croce CM, Huebner K. Lung cancer susceptibility in Fhit- deficient mice is increased by Vhl haploinsufficiency. Cancer Res. 65: 6576-82 (2005).
103. Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP. Epigenetic inactivatio of a RAS association domain family protein from the lung tumour suppressor locus 3p21.3. Nat Genet. 2000;25:315-9.
104. Serrano M, Gomez-Lahoz E, DePinho RA, Beach D, Bar-Sagi D. Inhibition of ras- induced proliferation and cellular transformation by p16INK4. Science. 1995;267: 249-52.
133 105. Byrd JC, Marcucci G, Parthun MR, Xiao JJ, Klisovic RB, Moran M, Lin TS, Liu S, Sklenar AR, Davis ME, Lucas DM, Fischer B et al. A phase 1 and pharmacodynamic study of depsipeptide (FK228) in chronic lymphocytic leukemia and acute myeloid leukemia. Blood. 105: 959-67 (2005).
106. Mund C, Hackanson B, Stresemann C, Lubbert M, Lyko F. Characterization of DNA demethylation effects induced by 5-Aza-2'-deoxycytidine in patients with myelodysplastic syndrome. Cancer Res. 65: 7086-90 (2005).
107. Aparicio A, Weber JS. Review of the clinical experience with 5-azacytidine and 5- aza-2'-deoxycytidine in solid tumors. Curr Opin Investig Drugs. 3: 627-33 (2002).
108. Yamada Y, Jackson-Grusby L, Linhart H, Meissner A, Eden A, Lin H, Jaenisch R. Opposing effects of DNA hypomethylation on intestinal and liver carcinogenesis. Proc Natl Acad Sci U S A. 102:13580-5 (2005).
109. Issa JP, Garcia-Manero G, Giles FJ, Mannari R, Thomas D, Faderl S, Bayar E, Lyons J, Rosenfeld CS, Cortes J, Kantarjian HM. Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent 5-aza-2'-deoxycytidine (decitabine) in hematopoietic malignancies. Blood. 103: 1635-40 (2004).
110. Lyko F, Brown R. DNA Methyltransferase Inhibitors and the Development of Epigenetic Cancer Therapies. J Natl Cancer Inst. 97: 1498-1506 (2005).
111. Baselga J, Norton L. Focus on breast cancer. Cancer Cell 8: 319-22 (2002).
112. Anttinen J, Kautiainen H, Kuopio T. Role of mammography screening as a predictor of survival in postmenopausal breast cancer patients. Br J Cancer. 94, 147-51 (2006).
113. Berry DA, Cronin KA, Plevritis SK, Fryback DG, Clarke L, Zelen M, Mandelblatt JS, Yakovlev AY, Habbema JD, Feuer EJ et al. Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med. 353: 1784-92 (2005).
114. EBCTCG. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet. 365: 1687- 717 (2005).
134 115. Ali S, Coombes RC. Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer. 2: 101-12 (2002).
116. Come SE, Buzdar AU, Ingle JN, Arteaga CL, Brown M, Dowsett M, Hilsenbeck SG, Kumar R, Johnston SR, Lee AV et al. Proceedings of the Fifth International Conference on Recent Advances and Future Directions in Endocrine Therapy for Breast Cancer: conference summary statement. Clin Cancer Res. 12: 997-1000s (2006).
117. Normanno N, Di Maio M, De Maio E, De Luca A, de Matteis A, Giordano A, Perrone F, et al. (2005). Mechanisms of endocrine resistance and novel therapeutic strategies in breast cancer. Endocr Relat Cancer. 12: 721-47 (2005).
118. Kurokawa H, Arteaga CL. ErbB (HER) receptors can abrogate antiestrogen action in human breast cancer by multiple signaling mechanisms. Clin Cancer Res. 9: 511-5 (2003).
119. Shou J, Massarweh S, Osborne CK, Wakeling AE, Ali S, Weiss H, Schiff R. (2004). Mechanisms of tamoxifen resistance: increased estrogen receptor-HER2/neu cross-talk in ER/HER2-positive breast cancer. J Natl Cancer Inst. 96: 926-35 (2004).
120. Shang Y, Brown M. Molecular determinants for the tissue specificity of SERMs. Science. 295: 2465-8 (2002).
121. Zwijsen RM, Buckle RS, Hijmans EM, Loomans CJ, Bernards R. Ligand-independent recruitment of steroid receptor coactivators to estrogen receptor by cyclin D1. Genes Dev. 12: 3488-98 (1998).
122. Jirstrom K, Stendahl M, Ryden L, Kronblad A, Bendahl PO, Stal O, Landberg G. Adverse effect of adjuvant tamoxifen in premenopausal breast cancer with cyclin D1 gene amplification. Cancer Res. 65: 8009-16 (2005).
123. Michalides R, Griekspoor A, Balkenende A, Verwoerd D, Janssen L, Jalink K, Floore A, Velds A, van't Veer L, Neefjes J. Tamoxifen resistance by a conformational arrest of the estrogen receptor alpha after PKA activation in breast cancer. Cancer Cell. 5: 597-605 (2004).
124. Hu H, Columbus J, Zhang Y, Wu D, Lian L , Yang S, Goodwin J, Luczak C, Carter M, Chen L et al. A map of WW domain family interactions. Proteomics. 4: 643-55 (2004).
135 125. Fan M, Long X, Bailey JA, Reed CA, Osborne E, Gize EA, Kirk EA, Bigsby RM, Nephew KP. The activating enzyme of NEDD8 inhibits steroid receptor function. Mol Endocrinol 16: 315-330 (2002).
126. Stewart SA, Dykxhoorn DM, Palliser D, Mizuno H, Yu EY, An DS, Sabatini DM, Chen IS, Hahn WC, Sharp PA et al. Lentivirus-delivered stable gene silencing by RNAi in primary cells. RNA. 9: 493-501 (2003).
127. Dutta K, Carmody MW, Cala SE, Davidoff AJ. Depressed PKA activity contributes to impaired SERCA function and is linked to the pathogenesis of glucose-induced cardiomyopathy. J Mol Cell Cardiol. 34: 985-96 (2002).
128. Ryan PD, Kopans DB, Sgroi DC. Case records of the Massachusetts General Hospital. Case 24-2005. A 58-year-old woman with early-stage estrogen-receptor- positive breast cancer. N Engl J Med. 353, 617-22 (2005).
129. Ropero S, Menendez JA, Vazquez-Martin A, Montero S, Cortes-Funes H, Colomer, R. Trastuzumab plus tamoxifen: anti-proliferative and molecular interactions in breast carcinoma. Breast Cancer Res Treat. 86: 125-37 (2004).
130. Bosher JM, Williams T, Hurst HC.. The developmentally regulated transcription factor AP-2 is involved in c-erbB-2 overexpression in human mammary carcinoma. Proc Natl Acad Sci U S A. 92: 744-7 (1995).
131. Bosher JM, Totty NF, Hsuan JJ, Williams T, Hurst HC. A family of AP-2 proteins regulates c-erbB-2 expression in mammary carcinoma. Oncogene. 13: 1701-7 (1996).
132. Turner BC, Zhang J, Gumbs AA, Maher MG, Kaplan L, Carter D, Glazer PM, Hurst HC, Haffty BG, Williams T. Expression of AP-2 transcription factors in human breast cancer correlates with the regulation of multiple growth factor signalling pathways. Cancer Res. 58: 5466-72 (1998).
133. Solberg R, Sandberg M, Natarajan V, Torjesen PA, Hansson V, Jahnsen T, Tasken K. The human gene for the regulatory subunit RI alpha of cyclic adenosine 3', 5'- monophosphate-dependent protein kinase: two distinct promoters provide differential regulation of alternately spliced messenger ribonucleic acids. Endocrinology. 138, 169-81 (1997).
136 134. Creighton CJ, Cordero KE, Larios JM, Miller RS, Johnson MD, Chinnaiyan AM, Lippman ME, Rae JM. Genes regulated by estrogen in breast tumor cells in vitro are similarly regulated in vivo in tumor xenografts and human breast tumors. Genome Biol 7: R28 (2006).
135. Jordan VC. Selective estrogen receptor modulation: concept and consequences in cancer. Cancer Cell 5: 207-13 (2004).
136. Fan M, Yan PS, Hartman-Frey C, Chen L, Paik H, Abbosh P, Cheng ASL, Li L, Huang TH-M, Nephew KP. Diverse gene expression and DNA methylation profiles correlate with differential adaptation of breast cancer cells to endocrine resistance (submitted).
137. Pellikainen J, Naukkarinen A, Ropponen K, Rummukainen J, Kataj V, Kellokoski J, Eskelinen M, Kosma VM. Expression of HER2 and its association with AP-2 in breast cancer. Eur J Cancer. 40: 1485-95 (2004).
138. Gee JM, Robertson JF, Gutteridge E, Ellis IO, Pinder SE, Rubini M, Nicholson RI. Epidermal growth factor receptor/HER2/insulin-like growth factor receptor signalling and oestrogen receptor activity in clinical breast cancer. Endocr Relat Cancer. 1: s99-111 (2005).
139. Rayala SK, Talukder AH, Balasenthil S, Tharakan R, Barnes CJ, Wang RA, Aldaz M, Khan S, Kumar R. P21-activated kinase 1 regulation of estrogen receptor-alpha activation involves serine 305 activation linked with serine 118 phosphorylation. Cancer Res. 66: 1694-701 (2006).
140. Yang X, Phillips DL, Ferguson AT, Nelson WG, Herman JG, Davidson NE. Synergistic activation of functional estrogen receptor (ER)- alpha by DNA methyltransferase and histone deacetylase inhibition in human ER-negative breast cancer cells. Cancer Res. 61: 7025-9 (2001).
141. Tang B, Peng ZH, Jiang J. The synergistic inhibitory effect of 5-aza-2-deoxycytidine and Tamoxifen on estrogen receptor alpha negative breast cancer cell lines in vitro. Zhonghua Wai Ke Za Zhi. 43: 1545-9 (2005).
137 142. Reid G, Metivier R, Lin CY, Denger S, Ibberson D, Ivacevic T, Brand H, Benes V, Liu ET, Gannon F. (2005). Multiple mechanisms induce transcriptional silencing of a subset of genes, including oestrogen receptor alpha, in response to deacetylase inhibition by valproic acid and trichostatin A. Oncogene 24: 4894-907 (2005).
143. Reid JF, Lusa L, De Cecco L, Coradini D, Veneroni S, Daidone MG, Gariboldi M, Pierotti MA. Limits of predictive models using microarray data for breast cancer clinical treatment outcome. J Natl Cancer Inst. 97: 927-30 (2005).
144. Jansen MP, Foekens JA, van Staveren IL, Dirkzwager-Kiel MM, Ritstier K, Look MP, Meijer-van Gelder ME, Sieuwerts AM, Portengen H, Dorssers LC, et al. Molecular classification of tamoxifen-resistant breast carcinomas by gene expression profiling. J Clin Oncol. 23: 732-40 (2005).
145. Ma XJ, Wang Z, Ryan PD, Isakoff SJ, Barmettler A, Fuller A, Muir B, Mohapatra G, Salunga R, Tuggle JT. A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell. 5: 607-16 (2004).
146. Benlimame N, He Q, Jie S, Xiao D, Xu YJ, Loignon M, Schlaepfer DD, Alaoui- Jamali MA. FAK signaling is critical for ErbB-2/ErbB-3 receptor cooperation for oncogenic transformation and invasion. J Cell Biol. 171: 505-16 (2005).
147. Howe LR, Chang SH, Tolle KC, Dillon R, Young LJ, Cardiff RD, Newman RA, Yang P, Thaler HT, Muller WJ. HER2/neu-induced mammary tumorigenesis and angiogenesis are reduced in cyclooxygenase-2 knockout mice. Cancer Res. 65: 10113-9 (2005).
148. Miller WR. Regulatory subunits of PKA and breast cancer. Ann N Y Acad Sci. 968: 37-48 (2002).
149. Reshkin SJ, Bellizzi A, Cardone RA, Tommasino M, Casavola V, Paradiso A. Paclitaxel induces apoptosis via protein kinase A- and p38 mitogen-activated protein- dependent inhibition of the Na+/H+ exchanger (NHE) NHE isoform 1 in human breast cancer cells. Clin Cancer Res. 9: 2366-73 (2003).
150. Shin I, Miller T, Arteaga CL. ErbB receptor signaling and therapeutic resistance to aromatase inhibitors. Clin Cancer Res. 12: 1008-1012s (2006).
138 151. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res. 59: 793-7 (1999).
152. Sanchez-Cespedes M, Esteller M, Wu L, Nawroz-Danish H, Yoo GH, Koch WM, Jen J, Herman JG, Sidransky D. Gene promoter hypermethylation in tumors and serum of head and neck cancer patients. Cancer Res. 60: 892-5 (2000).
153. Soria JC, Rodriguez M, Liu DD, Lee JJ, Hong WK, Mao L. Aberrant promoter methylation of multiple genes in bronchial brush samples from former cigarette smokers. Cancer Res. 62: 351-5 (2000).
154. Hennessy BT, Garcia-Manero G, Kantarjian HM, Giles FJ. DNA methylation in haematological malignancies: the role of decitabine. Expert Opin Investig Drugs. 12: 1985-93 (2003).
155. Blaese M, Blankenstein T, Brenner M, Cohen-Haguenauer O, Gansbacher B, Russell S, Sorrentino B, Velu T. Vectors in cancer therapy: how will they deliver? Cancer Gene Ther. 2: 291-7 (1995).
156. Zhang WW. Development and application of adenoviral vectors for gene therapy of cancer. Cancer Gene Ther. 6: 113-38 (1999).
157. Marini FC 3rd, Yu Q, Wickham T, Kovesdi I, Andreeff M. Adenovirus as a gene therapy vector for hematopoietic cells. Cancer Gene Ther. 7: 816-25 (2000).
158. Alemany R, Suzuki K, Curiel DT. Blood clearance rates of adenovirus type 5 in mice. J Gen Virol. 81: 2605-2609 (2000).
159. Tao N, Gao GP, Parr M et al. Sequestration of adenoviral vector by Kupffer cells leads to a nonlinear dose response of transduction in liver. Mol Ther. 3: 28-35 (2001).
160. Vlachaki MT, Hernandez-Garcia A, Ittmann M et al. Impact of preimmunization on adenoviral vector expression and toxicity in a subcutaneous mouse cancer model. Mol Ther. 6: 342-348 (2002).
139 161. Cantor JP*, Iliopoulos D*, Rao AS, Druck T, Semba S, Han SY, McCorkell KA, Lakshman TV, Collins JE, Wachsberger P, Friedberg JS, Huebner K. Epigenetic modulation of tumor suppressor expression in lung cancer xenografts suppresses tumorigenicity Int J Cancer (in press). *contributed equally.
162. Feng SS. Nanoparticles of biodegradable polymers for new-concept chemotherapy. Expert Rev Med Devices. 1: 115-25 (2004).
163. Ferrari M. Cancer nanotechnology: opportunities and challenges. Nat Rev Cancer. 5: 161-71 (2005).
140