FOXP3 IS A NOVEL X-LINKED BREAST SUPPRESSOR

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Tao Zuo, B.M.,M.s.

The Ohio State University 2006

Dissertation Committee:

Professor Yang Liu, Ph.D., advisor Approved by:

Professor Pan Zheng, M.D., Ph.D.

Professor Tim H-M. Huang, Ph.D.

Professor Michael Ostrowski, Ph.D Advisor

Molecular, Cellular and Developmental

Biology Graduate Program i ABSTRACT

Although BRCA1 and BRCA2 have been identified as the prototypes of the breast cancer suppressor , the genetic defects responsible for the majority of breast remain to be revealed. Cancer cells silence tumor suppressor genes (TSGs) by the two-hit mechanism in which loss-of-function mutations and

loss-of-heterozygousity (LOH) most frequently occur at the TSG loci. X-linked

tumor suppressors are of great interest since one allele of these genes can be

silenced by X- inactivation in females. The X-linked gene FoxP3 is a

member of the forkhead-box/winged-helix factor family. Germ-line

mutations of FoxP3 in both male mice (Foxp3sf) and man result in severe

with early lethality. Here we report that Foxp3sf/+ heterozygous mice develop malignancies at a high rate, the majority being mammary carcinomas that have inactivated the wild-type Foxp3 allele by skewed

X-inactivation while overexpressed two critical oncogenes, HER2/ErbB2 and

Skp2. We also show that FoxP3 represses these two oncogenes and inhibits tumor growth. Furthermore, FOXP3 is expressed in normal human breast epithelial cells but is silenced in most human breast cancers. The widespread gene deletion and somatic mutations at the FOXP3 gene were demonstrated by screening a large panel of human breast cancer samples. To

ii our knowledge, FOXP3 is the first X-linked breast cancer suppressor gene to be identified, and our data suggests that an interesting interplay between FOXP3 and oncogenes is involved in breast tumorigenesis.

iii

Dedicated to my family

iv ACKNOWLEDGMENTS

I would like to express my most heartfelt gratitude to all of the individuals

who supported me in many different ways during my 5-year graduate study at

OSU. First I would like to thank my advisor, Professor Yang Liu, whose guidance

and insight in science have been a great source of inspiration for my Ph.D.

training. I also would like to offer my sincere thanks to the members of my

graduate committee: Drs. Pan Zheng, Tim Huang, Mike Ostrowski and Charis

Eng, all of whom I have benefited scientifically. I deeply appreciate the

collaborative effort by Drs. Lizhong Wang, Carl Morrison, Xing Chang, Huiming

Zhang, Michael Chan and Richard Love. Also I sincerely value the input from all

the members in Drs. Liu/Zheng’s lab as well as the Core Facility in the

Department of Pathology. Additionally, to the many educators who have shaped

me over my long-term training journey in medicine and biomedical research, thank you!

Finally and most importantly, I’d like to thank my family. Especially my parents, as well as those relatives and close friends whom I have not been able to see for the last five years because of my dedication to this research project. I could not have gone through this tough journey at the other end of earth without their love and support from China. My dearest Aunt Lihua Zuo (左丽华) passed

v away last year. I was so grief-stricken for not being able to attend her funeral and mourn with my family at that painful moment. The only thing I can do here is to say: God’s peace to you my Aunt. She died of polycystic , a genetic disease. Her death reminds me that there is a long way to go before we can finally defeat these deleterious genetic diseases including cancer, and this is why we, as medical researchers, have to fight them at every moment with every piece of scientific progress in hand.

vi VITA

August, 17 1972…………………………….....Born in Kunming, Yunnan Province, P.R.China

1989 - 1994…………………………………….Undergraduate education in clinical medicine Peking University, Health Science Center (formerly Beijing Medical University)

1994……………………………………………….B.M. (Bachelor of Medicine) Peking University, Health Science Center (formerly Beijing Medical University)

1994-1997………………………………...... Surgical residency, Beijing Hospital Ministry of Health P.R.China

1997-2000………………………………...... Graduate Research Associate Institute of Geriatrics and Beijing Hospital; Peking Union Medical College. Ministry of Health, P.R.China

2000……………………………………………….M.s. in Medical Science, Institute of Geriatrics and Beijing Hospital; Ministry of Health P.R.China

2000-2001………………………………...... Urological specialist, Beijing Hospital Ministry of Health P.R.China

vii 2001-present……………………………………..Graduate Research Associate, Molecular, Cellular and Developmental Biology Graduate Program; Ohio State University

PUBLICATIONS

1. Yu Pulin, Zheng Hong, Su Hongxue, Zuo Tao, Gao Fankun,Wang Jianye. Prevalence of Prostatic Hyperplasia and its relative factors in six cities of China in 1997. Chinese Journal of Epidemiology 2000; 21:276-279

2. Zuo Tao, Wang Jianye. Apoptosis plays the critical role in development of Benign Prostate Hyperplasia. Chinese Journal of Urology. 1999; 20:508-510.

3. Wang Jianye, Zuo Tao, Wan Ben. The regulatory effect of TGF-beta on the proliferation and apoptosis of human prostatic epithelial cells. Chinese Journal of Urology. 2001; 22:229-232

4. Lizhong Wang, Tomonori Habuchi, Takeshi Takeahashi,Toshiyuki Kamoto, Tao Zuo, Kenji Mrrsumori, Norihiko Tsuchiya, Kazunari Sato, Tetsuro Kato. No association between Her-2 gene polymorphism at codon 655 and a risk of bladder cancer. Int. J. Cancer 2002; 97:787-790

5. Liu Xingluo, Gao JX, Wen J, Yin L, Li Ou, Zuo Tao, Gajewski T, Fu YX, Zheng P, Liu Y. B7DC/PDL2 Promeotes Tumor by a PD-1- independent Mechanism. J. Exp. Med. 2003, 197:1721-1730

viii 6. Xing Chang, Jian Xin Gao, Qi Jiang, Jing Wen, Nick Seifers, Lishan Su, Virginia L, Godfrey, Tao Zuo, Pan Zheng and Yang Liu The Scurfy mutation of FoxP3 in the stroma leads to defective thymopoiesis. J. Exp. Med. 005, 202:1141-1151.

7. Tao Zuo, Lizhong Wang, Carl Morrison, Xing Chang, Huiming Zhang, Yan Liu, Yin Wang, Xingluo Liu, Michael W.Y. Chan, Jin-Qing Liu, Richard Love, Virginia Godfrey, Rulong Shen, Tim H-M. Huang, Tianyu Yang, Pan Zheng and Yang Liu. FOXP3 is an X-linked breast cancer suppressor gene and an important repressor of HER-2/ErbB2 oncogene. Submitted (being peer- reviewed).

FIELDS OF STUDY

Major Field: Molecular, Cellular and Developmental Biology

ix

TABLE OF CONTENTS

Page Abstract...... …………………………………………………………………………...ii Dedication...... …………………………………………………………………….....iv Acknowledgments...... ……………………………………………………………….v Vita...... …...…………………………………………………………………...... vii List of Tables...... …………………………………………………………………xii List of Figures...... ……………………………………………………………..xiii

Chapters:

1. Introduction...……………………………………………………………...... 1

1.1 Tumor suppressor genes and oncogenes….………………………...1 1.2 and cancer...... ……………………...………...... 5 1.3 Forkhead transcription factors and cancer...... 8

2. Genetic evidence for FOXP3 as a novel X linked tumor suppressor Gene...……….……...…………………………………………………...... 12

2.1 Abstract...... ……………………………………………………….12 2.2 Introduction...... ……………………...... ……..13 2.3 Materials and Methods...... ……………………………………...14 2.4 Results...... ………………………………………………………..19 2.5 Discussion...... 27

3. FOXP3 suppresses breast cancer by repressing HER-2/ErbB2 oncogene…………..…………………………………………………………...56

3.1 Abstract...... ……………………………………………………….56 3.2 Introduction...... ……………………………………………...…...57 3.3 Materials and Methods...... ……………………………………...59 3.4 Results...... …...... 61 3.5 Discussion...... ……………………………………………………66

x

4. FOXP3 is a novel transcription suppressor for oncogene SKP2 in breast cancer……………………….……….………...... 83

4.1 Abstract...... …………………………………………...... 83 4.2 Introduction………………...... …………...... 84 4.3 Materials and Methods…………………………………...... …...86 4.4 Results……………………………………………………...... …..87 4.5 Discussion……………...... ……92

5. Concluding Remarks...... ……...... 107

Bibliography……………………………………………………...... ………………110

xi LIST OF TABLES

Table Page

2.1 Histological features of spontaneous cancers in wild-type and Foxp3sf mutant BALB/c mice………..……………………………………………..…..36

2.2 FOXP3 deletions in 28 cases of breast cancer.……………………………50

2.3 Summary of somatic FOXP3 mutations in breast cancer….……………...53

3.1 ErbB2 expression in mouse mammary tumors…..……………….………..69

3.2 Effects of FOXP3 deletions on HER2 expression in breast cancers……..………………………………….……………………81

3.3 Effects of FOXP3 mutations on HER2 expression in breast cancers….…………………..……………………….………………81

3.4 HER2/ErbB2 promoter region ChIP real-time PCR primers………..….....82

4.1 Skp2 Expression in mouse mammary tumors…..….……..………………..95

xii LIST OF FIGURES

Figure Page

2.1 Increased susceptibility to malignancies in heterozygous Foxp3sf/+ mutant mice………………………...…………………………………………..34

2.2 Histopathology of spontaneous tumors from heterozygous Foxp3sf/+ mice ………...……………………………………………...... 37

2.3 Histopathology of mammary glands and tumors in mice.…..………...…..39

2.4 No Otc expression detected in mammary gland ...……………………...…41

2.5 Foxp3 is silenced in mammary cancer by skewed X-inactivion...….....….42

2.6 Foxp3 inhibits the growth and tumorigenicity of multiple breast cancer cell lines……………………………………………………………...... 44

2.7 Human breast cancers have lost FOXP3 expression……………………...46

2.8 Deletion of FOXP3 locus in human breast cancer cells………………...... 48

2.9 Somatic mutations of FOXP3 in the breast cancer samples……………...51

3.1 Mammary tumors in Foxp3sf/+ mutant mice overexpress HER2/ErbB2 oncogene...……………...... …………………………………....70

3.2 FoxP3 is a transcription repressor of HER2/ErbB2 oncogene….…....…..72

3.3 Site-directed mutagenesis of Forkhead binding sites abrogates repression of HER2/ ErbB2 promoter by Foxp3…….…..………………....74

3.4 FOXP3 is an essential repressor of HER2/ ErbB2 in mammary epithelial cells………..…………………………...…...……………………....76

xiii 3.5 An inverse correlation between FOXP3 and HER2/ErbB2 among human breast cancer samples………………………….……..………….....78

3.6 FOXP3 suppresses HER2/ ErbB2 in human breast cancer……………....79

4.1 Inactivation of the FoxP3 locus results in increased SkP2 expression ………………...……………………………………………………….…...... 96

4.2 An inverse correlation between FOXP3 and SKP2 in human cancer samples………………………………………………………………………....98

4.3 FoxP3 directly represses Skp2 and indirectly up-regulates p27..…...... …99

4.4 FoxP3 is a transcriptional repressor for the Skp2 oncogene………...... 101

4.5 Inducing FOXP3 expression in Tet-off breast cancer cells causes cell death and reduces SKP2 expression…………………………..…...... 103

4.6 FoxP3 blocks cell cycle and causes a polyploidy in mammary cancer cells…..………………………………………….……....105

xiv

CHAPTER 1

INTRODUCTION

1.1 Tumor suppressor Genes and Oncogenes

The etiology of cancer is extremely complicated. However more and more compelling evidence demonstrates that cancer arises from genetic mutations in either germ cells or somatic cells, which cause hereditary malignancies or sporadic cancers (Alberts, 2002; Macdonald et al., 2004; Vogelstein and Kinzler,

2002). Some of these genetic abnormalities are gain-of-function mutations in the particular group of genes called proto-oncogenes. On the other hand, the loss-of- function mutations occurr in another group of genes called tumor suppressor genes (TSGs). By enhancing cell proliferation and survival, both kinds of mutations play a critical role in transforming normal cells into malignant tumors.

The breakthrough of cancer genetics in finding cancer related genes comes from studying viruses that are associated with malignant tumors. Transduction of Src oncogene by Rous sarcoma virus was first revealed to cause the tumor development 30 years ago (Stehelin et al., 1976). Subsequent studies discovered many other oncogenes transduced by viruses, such as v-ABL which causes , v-RAS related sarcoma and v-ERBB which was first shown to

1 result in erythro-leukemia but was later found to be a critical oncogene in breast

cancer (Alberts, 2002). The oncogene’s dominant transforming effect was

elegantly demonstrated by the tumor DNA transfection assay in which NIH-3T3 cells became to possess the traits of malignant cells through the transfection of

DNA derived directly from cancers (Shih et al., 1979). Other aberrant genetic

changes in the oncogenes in tumors include the amplification of genes such as

HER2/ErbB2 and , the chromosomal translocation of genes like ABL and

MYC, as well as the constitutively active mutations that are often found in RAS

oncogene (Alberts, 2002; Vogelstein and Kinzler, 2002).

On the other hand, mutations in tumor suppressor genes are generally

recessive in their effects on the individual cell, which means that both alleles of a

TSG have to be defective in order to show the physiology effects on the cells.

This is the reason why most of the prototype tumor suppressor genes have been

identified by studying the hereditary cancer syndromes that are usually

characterized by a germ-line mutation in the particular TSG. The heritable mutant

TSG passes on to the family members who are more susceptible to developing

cancers in their lifetime than a normal person is, because the chance of silencing

the TSG becomes higher only by disrupting the remaining allele. LOH (loss-of-

heterozygousity) was found to be the most frequently occurring mechanism by

which cancer cells finally nullify the TSG in their genome. By studying LOH in the

from cancer cells, RB gene was first cloned from familial

retinoblastoma patients (Cavenee et al., 1983; Dryja et al., 1986), which then

promoted the identification of other TSGs in different hereditary cancer

2 syndromes, such as BRCA1 and BRCA2 in familial breast cancer (Miki et al.,

1994; Wooster et al., 1995), and the APC gene in familial adenomatous polyposis coli (Groden et al., 1991; Joslyn et al., 1991). LOH is caused by several mechanisms, including the reduplication of the abnormal allele on the normal allele locus, an interstitial deletion of the normal chromosome, or a recombination event resulting in two copies of the deficient allele. More recently, the epigenetic methylation of CpG islands in TSG’s promoter region was established as another mechanism to silence TSGs in many cancers (Herman et al., 1994; Jones and Laird, 1999). After almost 20-years of extensive studies of different hereditary tumor syndromes, many TSGs have been identified, and among them has been found to be involved in more than half of the sporadic cancers (Alberts, 2002; Macdonald et al., 2004; Vogelstein and Kinzler, 2002).

Interestingly, many new TSG candidates do not conform with Knudson’s two-hit hypothesis (Knudson, 2001; Knudson, 1971). Some of them are called haploinsufficeint tumor suppressors as only a single allele is lost in the tumors

(Balmain, 2002; Macdonald et al., 2004; Vogelstein and Kinzler, 2002). For example, reduced levels of p27 were found in numerous cancers caused by LOH at a single allele while another remained intact (Pietenpol et al., 1995).

Hanahan and Weinberg recent postulated “six essential alterations in cell

physiology that collectively dictate malignant growth”, which include: self-

sufficiency in growth signals; insensitivity to growth inhibitory signals; evasion of

programmed cell death; limitless replication potential; sustained angiogenesis

and tissue invasion and metastasis (Hanahan and Weinberg, 2000). Given the

3 range of alterations in the biological functions during tumorigenesis, there is no doubt that many cancer related genes still remain unknown, especially recessive

TSGs that are not easy to identify compared to dominant oncogenes. Therefore

establishing new approaches to identify TSGs is imperative. Using an animal

model to study tumorigenesis is an attractive and practical way to search for

more unknown cancer related genes (Hennighausen, 2000; Venkatachalam and

Donehower, 1998), and the confirmation of the relevance to human cancers is

necessary in animal studies.

Although TSGs and oncogenes are genetically categorized in two different

groups, both of them are involved in the same physiological regulatory network in

cells by usually inputting either inhibitory (TSGs) or stimulatory (oncogenes)

effects on cell growth, survival and differentiation. Thus it is reasonable to imagine that some of the TSGs and proto-oncogenes interact with each other to accomplish their physiological functions in cells. Some of these interactions have been demonstrated. For instance, tumor suppressor APC functions as an ubiquitin ligase to degrade the β-Catenin; ARF interacts with onco-protein MDM2 to decrease its ability to destabilize p53; Cyclin D phosphorylates RB to release the blockage in the cell cycle (Alberts, 2002; Macdonald et al., 2004; Vogelstein and Kinzler, 2002). Most of the interactions identified so far are limited to the post-translation (protein) level, though several reports suggest the transcriptional level regulatory mechanism between TSGs and oncogenes. For example, tumor suppressor WT1 functions as a transcriptional repressor to decrease expression of BCL-2 (Mayo et al., 1999). Since gene transcription is one of the most

4 important steps in gene regulation, one would not be surprised to find that there

are far more undefined interactions between TSGs and oncogenes at the

transcriptional level.

1.2 X chromosome and cancer

The epigenetic mechanism of X inactivation

In mammals, females carry two X chromosomes while males have only

one. A dosage compensation of X-linked genes in females is achieved by X

inactivation, in which one of the two X chromosomes in the female is subject to

transcriptional silencing during the early stage of embryonic development (Boumil and Lee, 2001; Heard, 2004). Although the detailed process of X inactivation remains elusive, epigenetical mechanisms underlying X inactivation have been revealed at a rapid pace since Mary Lyon proposed this hypothesis 45 years ago

(Boumil and Lee, 2001; Heard, 2004; Huynh and Lee, 2003; Lee, 2005; Lyon,

1961; Mak et al., 2004; Okamoto et al., 2004; Xu et al., 2006). In the early 90s,

Hunt Willard’s group identified an X-linked noncoding RNA whose expression was restricted to the X chromosome that would later be inactivated (Brown et al.,

1991). Named XIST (X inactivation-specific transcript), this X linked gene encodes 15 to 17 kb transcript that accumulates and covers on the inactivate X chromosome. XIST has been shown to initiate but not maintain X inactivation

(Boumil and Lee, 2001; Heard, 2004). About 10 years ago Jeannie Lee and her colleagues first identified XIC (X-inactivation center) on X chromosome at Xq13.3 that contains XIST and spans around 40 to 80 kb region (Lee et al., 1996). Since

5 then accumulation of data has shown that XIC is involved in three initial steps of

X inactivation, namely counting, choice and initiation of silencing X chromosome.

Three loci in XIC are believed to be critical in regulating XIST expression therewith initiation of X inactivation; they are Tsix, DXPas34 and Xite (X- inactivation intergenic transcription elements) (Boumil and Lee, 2001; Heard,

2004; Lee, 2005). Tsix is an antisense locus of XIST and represses XIST expression. The X chromosome that expresses more Tsix instead of XIST is not subjected to inactivation, so Tsix regulates X-chromosome choice (Lee et al.,

1999). DXPas34 is a CpG-rich minisatellite marker that controls expression of

XIST by different methylation statuses on activate or inactivate X chromosome

(Courtier et al., 1995). Xite is also a noncoding gene and has recently been shown to be responsible for X-chromosome counting to ensure inactivation of all but one X chromosome in the cells (Lee, 2005). After XIST induces silencing of X chromosome during ES (embryonic stem) cell differentiation, the X inactivation status is irreversible. To maintain X inactivation, chromatin modifications such as hypoacetylation of histones H3 and H4, and hypermethylation of H3K9 and

H3K27 must take place (Heard, 2004). The histone methyltransferase

Eed/Enx1complex has been shown to interact with X chromosome shortly after

XIST RNA coating and cause methylation of H3K27(Heard, 2004). Also the histone variant macroH2A was found to incorporate Xi (inactivated X chromosome) (Boumil and Lee, 2001). Together with DNA methylation, the inactive state of Xi could be sufficiently maintained by the modifications listed above.

6 The very exciting progress in the field is that the reactivation of Xp

(paternal X Chromosome) from an imprinted status was found during the very

early embryonic stage, which ranges from the two- or four-cell stage to the

blastocyst stage (Heard, 2004). Only ICM (inner cell mass) is able to reverse this

Xp imprinting but extraembryonic tissues remains imprinted status on Xp. As a

result, Xp in ICM is capable of processing random X inactivation in the later

developmental stages (Huynh and Lee, 2003). Because pluripotent cells in ICM

subsequently develop to the fetus, the mosaic pattern of inactivation either Xp or

Xm (maternal X chromosome) exists in adult cells. This epigenetic plasticity

shown by the erasure of imprinted Xp during early embryonic development raises

more interesting questions regarding the possibility of epigenetically

reprogramming the genome in pluripotent stem cells or even in cancer stem cells.

X chromosome inactivation and cancer

X chromosome is one of the most interesting topics in cancer genetics

because only one X-chromosome, unlike autosomes, remains activated in

females and males, however relatively few works have been done in this field so

far (Spatz et al., 2004). David Livingston’s group had the most significant

contribution in 2002, which reported that BRCA1 plays a critical role in X

inactivation and indicated that defects in X inactivation might be involved in

breast cancer genesis (Ganesan et al., 2002). Most recently, the same group

(Richardson et al., 2006) reported that severe X chromosomal abnormalities in

one type of breast cancer, called BLC (basal-like cancer), include most of the

BRCA1 associated breast cancer but also some non-BRCA-related sporadic

7 cancers. BLCs are characterized as high grade invasive ductal carcinomas

expressing specific basal breast epithelial cytokeratins with negative expression

of HER2/Neu and ER (estrogen ), as well as highly aneuploid and

mutated p53. Interestingly, a series of Norwegian epidemiological studies

reported the high frequency of skewed X inactivation in young breast cancer

patients with non-BRCA1/BRCA2 familial breast cancer (Kristiansen et al., 2005;

Kristiansen et al., 2002). No significant association between skewed X inactivation and BRCA1/BRCA2 mutations in these studies might be caused by limited samples which only included 35 patients with BRCA1/BRCA2 hereditary mutations. Another possible explanation is that other unidentified X linked tumor suppressor genes cause the skewed X inactivation in breast cancer patients. The surprising fact in the cancer genetic field is that no putative X-linked breast cancer suppressor genes have been identified yet (Spatz et al., 2004). Therefore the relationship between these X-linked genes and cancer genesis is an open field to explore.

1.3 Forkhead transcription factors and cancer

The forkhead transcription factors were named after the Forkhead (Fkh) gene that was identified and found to be the causative mutated gene in the fruit fly fork head mutant that has dual-head structures in embryos (Weigel et al.,

1989). The hepatocyte nuclear factor 3 was the first mammal gene to be found to share a highly conservative DNA binding domain with the Forkhead gene (Lai et al., 1991). This 110-amino-acid DNA binding domain belongs to the helix-turn-

8 helix domain family but has its unique butterfly wing like structure characterized

by 3 alpha-helix core flanked with two large loops or wings, so that this group of transcription factors is also named winged helix protein (Carlsson and Mahlapuu,

2002). During the last decade, an increasing number of forkhead transcription factors have been identified in different species. In 2000, the new nomenclature system was induced to unify inconstant names and classifications in the field

(Kaestner et al., 2000). The 17 subfamilies, named from FOXA to FOXQ, of forkhead were categorized by phylogenetic analysis of their amino acid sequence. Since the amino acid sequence is well conserved in the forkhead domain, most fork transcription factors were found to bind a seven-nucleotide core consensus DNA motif: RYMAAYA (R=A or G; Y=C or T; M=A or C)

(Carlsson and Mahlapuu, 2002). Although most of the forkhead transcription factors contain the transcription activation domain, some new identified members are transcription repressors that include all 4 forkhead members in the FoxP subfamily (Li et al., 2004; Shu et al., 2001; Wang et al., 2003; Ziegler, 2006). The conserved forkhead, and domains are found in all FoxP proteins, but Foxp3 misses most of the N terminal parts that contain poly- glutamine domain shared by the other 3 FoxP proteins. The zinc finger and leucine zipper domains have been shown to mediate the dimerization between

FoxP proteins, which was rarely found in other forkhead proteins that usually bind DNA as monomers (Li et al., 2004; Schubert et al., 2001; Wang et al., 2003;

Ziegler, 2006).

9 The forkhead proteins are involved in diverse biological functions, such as

embryonic development, metabolic balance, cell cycle, cell survival and immunity

(Burgering and Kops, 2002; Carlsson and Mahlapuu, 2002; Coffer and Burgering,

2004; Czech, 2003; Katoh, 2004; Ramsdell and Ziegler, 2003; Ziegler, 2006).

Lines of evidence also point out that forkhead proteins play a critical role in

tumorigenesis (Banham et al., 2001; Fox et al., 2004; Katoh, 2004). The most

widely studied cancer related forkhead proteins are from the FoxO subfamily that

includes FoxO1, 3, 4. All of these three members are involved in chromosomal

translocation in the different malignancies (Katoh, 2004). FOXO1 was originally

named FKHR (Forkhead in rhabdomyosarcoma) because some alveolar

rhabdomyosarcoma patients are found to carry the chromosomal translocations

t(2;13) or t(1;13) that cause the trans-activation domain of FOXO1 to fuse to the

DNA binding domain of PAX3 and PAX7 respectively (Xia et al., 2002). The

fused PAX3-FKHR or PAX7-FKHR proteins gain the stronger transcriptional

activity and function as oncogenic transcription factors for PAX3 and PAX7

targets. Trans-activation domains of FOXO3 and FOXO4 were also found to fuse

with the DNA binding domain of MLL through chromosomal translocations in

some acute lymphoid leukemia patients (So and Cleary, 2004). These

abnormally fused transcription activators accelerate tumor growth not only by their enhanced aberrant transcription activity but also through the disruption of

normal function of FOXO proteins that usually repress cell cycle by activating p27

and induce proapoptotic genes expression (Accili and Arden, 2004; Burgering

and Kops, 2002). Most recently, FOXO1 was found as a new target of the

10 oncogenic protein SKP2, which is the F-box protein in the SCF complex (SKP1-

CUL1-F-box-protein) and induces the degradation of target proteins through

ubiquitin-proteasome pathway (Huang et al., 2005; Nakayama and Nakayama,

2006). Although FOXO proteins possess the tumor suppressor functions and are

involved in some tumor genesis, they have not been confirmed as tumor

suppressor genes. Recently, several reports (Banham et al., 2001; Fox et al.,

2004) have suggested that FOXP1 is a possible tumor suppressor gene by

showing decreased expression of FOXP1 in colon and breast cancers. Though

FOXP1 locates in chromosome 3p14.1, a region commonly showing loss of heterozygosity in many tumors, more relevant genetic evidence such as mutations or epigenetic inactivation of this gene in cancer is not at hand yet. On the contrary, FOXG1 is considered a proto-oncogene because it is the mammalian ortholog of avian retroviral oncogene qin (Katoh, 2004; Li and Vogt,

1993). As forkhead proteins are intensively involved in many important physiological processes such as cell cycle, survival and metabolic balance that are so critical in the cell transformation and tumor growth, it is very tempting to ask whether and how this group of unique transcription factors connects to cancer.

11

CHAPTER 2

GENETIC EVIDENCE FOR FOXP3 AS A NOVEL X LINKED TUMOR SUPPRESSOR GENE

2.1 Abstract

Tumor suppressor genes (TSGs) are the most critical cancer related genes,

because loss-of-function of TSGs causes unrestricted cell growth and results

genesis of malignancy (Vogelstein and Kinzler, 2002). The most frequently

occurring genetic changes silencing TSGs in cancer cells are loss-of-function

mutations and LOH (loss of heterozygousity) (Knudson, 2001; Knudson, 1971;

Vogelstein and Kinzler, 2002). X linked TSGs are of great interest since one

allele of these genes can be silenced by X chromosome inactivation in females

(Spatz et al., 2004). The X linked FoxP3 is a member of the forkhead/winged

helix family proteins. Germ line mutations of FoxP3 in both

male mice (Foxp3sf) and man result in severe autoimmune disease with early

lethality (Bennett et al., 2001; Brunkow et al., 2001; Wildin et al., 2001). Here we

report that Foxp3sf/+ heterozygous mice develop malignancies at a high rate. The

majority of cancers are mammary carcinomas in which the wild-type Foxp3 allele is silenced by skewed X inactivation. We also show that FOXP3 is expressed in

normal human breast epithelial cells but silenced in most human breast cancers.

12 Furthermore, widespread gene deletion and somatic mutations at FOXP3 gene

locus were found in a large panel of human breast cancers. Taken together, we

have compelling genetic evidence to demonstrate that FOXP3 is a novel X linked

tumor suppressor gene.

2.2 Introduction

Cancer is a genetic disease that arises from genetic mutations in either

germ cells or somatic cells, which cause hereditary malignancies or sporadic

cancers (Alberts, 2002; Macdonald et al., 2004; Vogelstein and Kinzler, 2002).

Mutations in TSGs (tumor suppressor genes) are generally recessive in their effects on the individual cell, which means that both alleles of a TSG have to be

defective before its physiological function is altered in cells. Therefore a particular

TSG is usually silenced in cancer by at least two independent genetic changes at

its locus, which is referred as to “two-hit” theory (Knudson, 2001; Knudson, 1971).

Loss-of-function mutations and loss-of-heterozygousity (LOH) most frequently

occur at TSG loci in many types of malignancies (Knudson, 2001; Knudson, 1971;

Vogelstein and Kinzler, 2002). By following the “two-hit” mechanism described

above, people have identified a lot of putative TSGs, like RB, BRCA1 and

BRCA2 genes. However, a lot of unknown TSGs still remain to be revealed, for

instance, the genetic defects responsible for half familial breast cancers are still

unknown (Wooster and Weber, 2003).

FoxP3 is among the newest members of the forkhead winged helix family.

It was identified during position cloning of Scurfin, which is mutated in an X-linked

13 autoimmune disease mouse model called Scurfy (Brunkow et al., 2001). Germ-

line mutations of human FOXP3 were found in human IPEX (Immune

dysregulation, Polyendopathy, Enterophathy, X-linked) patients (Bennett et al.,

2001; Wildin et al., 2001). Surprisingly, in our analysis of the immune function of

mice heterozygous for the Foxp3 mutation, we observed a high rate of

spontaneous mammary cancer. We therefore systematically analyzed if the

Foxp3 gene is a novel breast tumor suppressor gene. First we show that FoxP3

was silenced in both mouse mammary cancer and most human breast cancers;

second, our data suggests that mammary cancer cells lost Foxp3 expression by skewed X-inactivation. Third, like most TSGs, FoxP3 dramatically repressed growth of cancer cells. Finally, we examined a large panel of human breast cancers and showed widespread gene deletion and somatic mutations of FOXP3.

Taken together, our data strongly suggest that FOXP3 is a novel breast cancer suppressor gene.

2.3 Materials and Methods

Cells, antibodies and animals

TSA is an aggressive and poorly immunogenic cell line established from the first in vivo transplant of a moderately differentiated mammary

adenocarcinoma that arose spontaneously in a BALB/c mouse (Giovarelli et al.,

1995). SKBR-3 (Fogh et al., 1977) and MCF7 (Soule et al., 1973) are breast

cancer cell lines purchased from ATCC (Manassas, VA, USA). Affinity purified

rabbit anti-Foxp3 antibodies, specific for Foxp3 N-terminus peptide starting at

14 position 25: (C)LLGTRGSGGPFQGRDLRSGAH, were provided by Dr. Lishan Su at the University of North Carolina (Chapel Hill, NC, USA). Other antibodies were purchased from the following vendors: Santa Cruz Biotechnology, Inc

(Santa Cruz, CA, USA), ERα (MC-20), PRα (C-19), ErbB2 (C-18), and normal rabbit IgG. Dako (Carpinteria, CA, USA), HER-2/NEU (A0485). Scurfy female mice with either two (Foxp3sf/+Otcspf/+) or one (Foxp3sf/+) mutations were

backcrossed to BALB/c background for at least 12 generations at the University

of North Carolina at Chapel Hill. Sparse-fur (Otcspf/+) female mice used for the

studies were obtained from the double mutant mice after a rare re-segregation of

the two mutations at the Ohio State University Laboratory Animal Facility.

Littermates of the mutants were used as controls. All mouse genotyping PCR

followed previously published protocols (Chang et al., 2005).

Histology, immunohistochemistry and OTC histochemistry

Tissues were fixed in 10% formalin overnight and embedded in paraffin.

For immunohistochemistry, tissue sections were incubated with the primary

antibodies. The biotinylated second-step antibodies were applied to tissue

sections and then incubated with the avidin biotinylated complex reagents

(Vector). Staining was developed by using the DAB peroxidase substrate kit

(Vector). Histochemical analysis of OTC activity was performed as previously

described before (Ye et al., 1996), and slides were counterstained with Nuclear

Fast Red (Vector).

15 Laser capture microdissection and RNA extraction

Frozen tissue was embedded in OCT and sectioned at –20° C. Sections

were fixed immediately in 75% DEPC ethanol for 30 seconds, then washed with

DEPC water for 15-30 seconds. The brief Hematoxylin/Eosin staining (10

seconds) was performed in RNase free condition. Dehydration of sections was

performed by sequentially incubating slides in 95%, 100% ethanol for 30 seconds

and then in xylene for 5 minutes. Air-dried sections were ready for LCM.

Microdissection of breast epithelial cells, breast cancer cells, hepatocytes and

in spleen were performed using an Arcturus PixCell II Laser Capture

microdissecting microscope (Arcturus Engineering Inc., Mountain View, CA).

RNA from microdissected tissue was extracted by using the PicoPure RNA Kit

and by following the manufacturer’s instruction (Arcturus Engineering Inc.,

Mountain View, CA).

Quantitative real-time PCR

Relative quantities of mRNA expression were analyzed using real-time

PCR (Applied Biosystems ABI Prism 7700 Sequence Detection System, Applied

Biosystems). The SYBR (Qiagen) green fluorescence dye was used in this study.

The primer sequences (5’-3’) were as follows: Foxp3

forward:ATCTCCTGGATGAGAAAGGCAAGG, and

reverse:TGTTGTGGAAGAACTCTGGGAAGG ; Hprt

forward:AGCCTAAGATGAGCGCAAGT and reverse:

TTACTAGGCAGATGGCCACA; murine cytokeratin 19, forward 5' -

ACCCTCCCGAGATTACAACC-3', reverse 5' -CAAGGCGTGTTCTGTCTCAA-3'.

16 murine cd3, forward 5' -TCTGCTGGATCCCAAACTCT-3', reverse 5' -

TGCACTCCTGCTGAATTTTG-3'. The data were collected and analyzed by the

ABI Prism 7700 sequence detection system software. Additional primers for ChIP assays are available upon request.

Plasmids, transfection and colony forming assay

The full-length mouse Foxp3 cDNA and Otc cDNA were cloned into the

pcDNA3 vector between the EcoRI and XbaI sites for transient transfection

experiments. Also Foxp3 was cloned into the pEF1/V5 vector (invitrogen) for

stable transfection experiments. The sequence of this insert has been confirmed.

Breast cancer cells were seeded to 50-60% confluence, and FuGene6 (Roche)

was used as carrier of the DNA for transfection by following the manufacturer’s

instruction. At 48 hours after transfection, the transfectants were selected with G-

418 (Invitrogen) to eliminate cancer cells that received no vectors. 3-4 weeks

later, the drug-resistant clones were stained with 1% violet crystal dye.

FISH (Fluorescence in situ hybridization) detecting Xp11.2 in breast cancer

FISH for FOXP3 deletion was carried out using BAC clone RP11-344O14

(ntLocus X: 48,817,975-48,968,223), which was verified by PCR to contain the

FOXP3 gene. The minimal common region of deletion was identified using

flanking p-telomeric and centromeric clones, RP11-573N21 (ntLocus X:

43,910,391-44,078,600) and RP11-353K22 (ntLocus X: 54,416,890-54,545,788),

respectively. Locus specific BAC clones were labeled with spectrum orange

using commercially available reagents per the manufacturer’s recommendations

(Vysis, Downers Grove, Ill.). Chromosome X enumeration was done by using a

17 commercially available spectrum green CEPX probe (Vysis, Downers Grove, Ill.).

Cutoff values for the determination of deletion of each probe were established by

scoring 200 nuclei from forty 0.6-millimeter cores representing normal tissue from

10 different organs. Cutoff values were then established by calculation of the mean plus three times the standard deviation of the number of normal cells with a false-positive signal. For BAC clones RP11-344O14, RP11-573N21, and

RP11-353K22 these numbers were 7.1%, 8.1%, and 8.0%, respectively, meaning only cases of breast cancer with greater than this percentage of cells

with one or two CEPX signals and none or a single locus specific signal,

respectively, were counted as abnormal. For all FISH done in this study a total of

at least 200 nuclei were scored for every case. For virtually all cases with FOXP3

deletion, the percentage of cells with reduced number of FOXP3 greatly

exceeded the cut-off value. These were thus considered clear-cut cases of gene

deletion.

All FISH were done using standard protocols optimized for breast cancer

specimens. Briefly, formalin fixed, paraffin-embedded tissue microarray blocks

were cut into 3 to 4 µm thick sections, incubated over night at 560C, deparaffinized, washed, digested with protease, formalin fixed, denatured, and hybridized at 37°C for 16 hours. The slides were then washed in a post- hybridization wash, counter stained with 4'-6-Diamidino-2-phenylindole (DAPI), and covered with a coverslip. Specimens were evaluated with an Olympus BX51 microscope (Olympus Optical Company, LTD., Japan) under oil immersion at x150 magnification using the recommended filters.

18 2.4 Results

2.4.1. Foxp3sf heterozygous mutant mice have a higher risk for

malignancies.

Before the identification of the scurfy gene as Foxp3, the scurfy (sf) locus had

been mapped in coupling and repulsion crosses to a location near the sparse fur

(spf) locus, which encodes for ornithine transcarbamylase (OTC) (DeMars et al.,

1976; Veres et al., 1987). The mutant BALB/c mice we used for the initial study

thus carried mutations in two closely linked X-chromosome genes, Foxp3sf and

Otcspf. During the course of the study, a spontaneous segregation of Otcspf allowed us to obtain a BALB/c Otcspf/+ strain. Meanwhile, we obtained an

independent line of Scurfy mice that has never been crossed to the Spf mutant

mice and backcrossed the Scurfy mutant allele (Foxp3sf) for more than 12

generations into the BALB/c background (Chang et al., 2005). Hemizygous male

mice, Foxp3sfOtcspf and Foxp3sf, died within 5 weeks of age, while female mice

with only one copy of FoxP3 gene survived to adulthood and appeared normal within the first year of life with a normal function (Fontenot et al., 2003;

Godfrey et al., 1994). These mice were used as breeders and were retired when

they reached about one year of age. Our extended observations of the retired

breeders for up to two years revealed that close to 90% of the Foxp3sf/+Otcspf/+

and Foxp3sf/+ mice spontaneously developed malignant tumors. More than

seventy percent of the tumors were mammary carcinomas (Fig. 2.1A), and

histological analyses revealed lung metastasis (Fig. 2.1A lower panels) in about

40% of the mice with mammary cancer (Table 2.1). More than a third of the

19 tumor-bearing mice had multiple lesions in the mammary gland. Most, although

not all, mammary carcinomas expressed (ER+, 14/18; Figure

2.3J) and (PR+, 12/18; Figure 2.3K). The cancer

incidences in the littermate controls and a line of congenic mice with a mutation in Otc but not Foxp3 were comparable with each other (Figure 2.1B).

In order to focus on mammary cancer, we treated the mice with a carcinogen, 7,12-dimethylbenz [a] anthracene (DMBA) in conjunction with progesterone, which resulted in an accelerated rate of mammary cancer in the

BALB/c mice (Aldaz et al., 1996; Lydon et al., 1999). Mice heterozygous for

Foxp3sf/+, but not those heterozygous for Otcspf showed substantially increased

susceptibility to mammary cancer, as revealed by earlier onset, increased

incidence (Figure 2.1C) and multiplicity of the breast tumors. These data

demonstrated that mutation of Foxp3, but not Otc resulted in a major increase in

susceptibility to mammary carcinoma.

2.4.2. Spontaneous tumor spectrum and mammary phenotype in Foxp3sf heterozygous mutant mice

Beside mammary cancers, there were 4 other types of histologically distinct malignancies in Foxp3sf/+ heterozygous mutant mice, namely lymphoma,

hepatoma, sarcoma and uterus adenocarcinoma (Table 2.1 and Figure 2.2).

Lymphoma, affecting around 10% of animals, was the second most frequent

tumor occurring in heterozygous Foxp3sf/+ mice. Both T cell lymphoma (6 cases)

and B cell lymphoma (2 cases) were observed (Figure 2.2A-F). In addition, 5

mice were found to develop both lymphomas and mammary tumors. The

20 lymphomas in Foxp3sf/+ heterozygous mutant mice had a very invasive

phenotype in which malignant lymphocytes invaded multiples organs, including

liver, lung and intestine (Figure 2.2D-F). Three liver adenomas and one liver

carcinoma (Figure 2.2 G, H) were found in 4 mice that also carried mammary

tumors. Sarcomas were detected in three mice (Figure 2.2J,K and 2.3C),

including 2 hemangiosarcomas and one leiomyosarcoma. Interestingly, two

sarcomas grew and aggressively invaded in mammary glands. Uterus was

another sexual organ in which spontaneously tumors arose (Figure 2.2I); and

ovarian cancers counted around half of all tumors observed in DMBA treated

Foxp3sfOtcspf mice (Data not shown). The fact that Foxp3sf/+ mutant mice develop tumors in these sexual organs suggests that Foxp3 may play an important role in physiological hormone regulation.

Mammary gland hyperplasia is usually found in most mouse mammary tumor models (Cardiff et al., 2000). We detected severe hyperplasia in all of the

12 Foxp3 mutant mice analyzed, but in only 2 out of 12 age-matched wild-type control female mice (Figure 2.3A,B). The hyperplastic lesions were usually multifocal in mammary glands and characterized by hyperchromatic ducts and lumen with more than one layer of atypical epithelia cells but intact basement membrane surrounded by thickened connective tissue in terminal ductal lobular units, which fits with the diagnosis of MIN (mammary intra-epithelial neoplasia)

(Cardiff et al., 2000). Although we were not able to find in situ carcinomas in

Foxp3sf/+ mutant mice, given the highly malignant potential of MIN, it was very

possible that in situ carcinoma was also involved in the early stages of

21 tumorigenesis in the Foxp3sf/+ mutant mice. Solid, papillary and squamous carcinomas were the three types of histological found in mammary

cancers developed from Foxp3sf/+ mutant mice (Figure 2.3D-I). In terms of

malignant phenotypes that included invasive growth patterns and lung metastases, we did not find any differences among the three types of mammary

cancers. However, this diverse histological phenotype might indicate that Foxp3

is involved in different pathways during mammary tumorigenesis in the mutant

mouse.

2.4.3. Foxp3 is silenced in mammary cancer by skewed X-inactivion

For Foxp3 to act as a tumor suppressor gene, it should be expressed in

normal mammary tissue and be inactivated in cancerous tissue. Since

expression of Foxp3 has not been reported in mammary tissue, we isolated

normal and cancerous cells by laser-capture microdissection and compared

expression of Foxp3 (Figure 2.5A) and Otc by real-time RT-PCR and histochemistry (Figure 2.4). The complete absence of the cd3 transcript (Figure

2.5A) indicated that the micro-dissected samples were devoid of T cells, the main cell types known to express Foxp3 (Fontenot et al., 2003; Ziegler, 2006). A representative profile and summary data of Foxp3 expression in Foxp3sf/+Otcspf/+ mice and age-matched WT control mice are shown in Figure 2.5B. Foxp3 mRNA was detected in normal mammary epitheliums from both the WT and

Foxp3sf/+Otcspf/+ mice, but not in mammary cancer cells from the same

Foxp3sf/+Otcspf/+ mice. Immunohistochemistry staining confirmed the loss of

22 expression of Foxp3 in the mammary carcinoma generated from the

Foxp3sf/+Otcspf/+ mice.

Foxp3 is an X linked gene that is subject to X chromosomal inactivation

(Carrel and Willard, 2005; Tommasini et al., 2002). The scurfy mutation is

caused by insertion of AA dinucleotide resulting in a frameshift and early termination codon. Previous studies have established that this mutation causes dramatic reduction of Foxp3 mRNA, most likely due premature termination

codon-mediated RNA decay. It is therefore possible that absence of Foxp3

expression in mammary cancer cells is due to X-chromosome inactivation of WT

alleles, leaving the mutant allele active but unable to accumulate significant level

of Foxp3 mRNA. To test this possibility, we carried out an anchored RT-PCR and

cloned the low level of Foxp3 mRNA in the breast tissues. We sequenced the

cDNA clones from pooled samples after ruling out potential T cell contamination

(based on lack of T-cell specific cd3 transcripts). As shown in Figure 2.5C, 100%

of the Foxp3 transcripts in the cancerous tissues were from the mutant alleles,

which indicated that the wild-type allele was silenced in the tumor cells. In

contrast, the transcript from the mutant allele constituted only 15% of the

transcript in the normal mammary samples from the same mice. The under-

representation of mutant allele in normal tissue likely reflects the reduced mRNA

stability of the mutant transcript rather than preferential silencing of the mutant

allele. Thus, the expression pattern of Foxp3 fulfills another criterion for a tumor

suppressor gene.

23 2.4.4. Foxp3 suppresses the growth of breast cancer cells in vivo and vitro

To test whether the Foxp3 gene could suppress the growth of breast cancer cells, we transfected empty vector, or vectors carrying either Foxp3 (mouse or

human origin) or Otc cDNA into three breast cancer cell lines, including the mouse mammary tumor cell line TSA, and the human breast cancer cell lines

MCF7 (ER+HER-2low, no HER-2 amplification), SKBr3 (ER-HER-2high with HER-2

amplification). Since all vectors carried a neomycin-resistance gene, the

untransfected cells were removed by a selection with G418. The numbers of drug

resistant cells were comparable among different groups initially, which suggest

comparable transfection efficiency. However, while the vector-transfected cells

grew rapidly, the Foxp3-transfected cell lines seldom grew into large colonies.

The Foxp3-transfected culture had drastic reduction in both the size and the

number of the drug-resistant colonies. No effect was observed when the Otc

cDNA was used (Figure 2.6A&B). Thus, Foxp3, but not Otc, suppresses the

growth of mouse and human breast cancer cell lines.

We transfected TSA cells with either empty vector or V5-tagged Foxp3

cDNA. The stable transfectant cell lines were selected by G-418. The vector and

Foxp3-V5-transfected cell lines were injected into syngeneic BALB/c mice, which

were then observed for tumor growth and mouse survival. As shown in Figure

2.6C, Foxp3-transfectants showed reduced tumor growth in vivo. The mice that

received TSA-Vector cells became moribund earlier with higher incidence, while

about 50% of the mice that received the Foxp3-V5-transfected cells survived

more than 7 weeks (Figure 2.6D).

24 2.4.5. Genetic evidence for FOXP3 as a tumor suppressor gene in human

breast cancer

We took three approaches to determine whether the findings in the mutant mice are relevant to the pathogenesis of human breast cancer. First, we used immunohistochemistry to determine expression of FOXP3 in normal vs. cancerous tissue. As shown in Figure 2.7A, while more than 80% of the normal breast samples expressed FOXP3 in the nuclei of the epithelial cells, less than

20% of the cancerous tissue showed nuclear staining. The difference between

normal and cancer tissue is highly significant (P=6.0x10-16, Chi-square test). The

reduced expression of FOXP3 was observed in both ER+ (P=1.0x10-7,Chi-square

test) and ER- (P=8.2x10-18,Chi-square test) cancer samples, although more

severe reduction was observed in the ER- than in the ER+ tumor group

(P=0.002,Chi-square test). As shown in figure 2.7B, RT-PCR revealed that the

FOXP3 transcripts in primary human mammary epithelial cells (HMEC), but much lower in most of the 9 human breast cancer cell lines tested. Also HMEC contained two major forms of FOXP3 which were found in human T cells, and the level was about 70-fold lower in the breast epithelial cells (Figure 2.7B lower panel).

Second, we used fluorescence in situ hybridization (FISH) to determine whether the FOXP3 gene was deleted in breast cancer samples. The minimal common region of deletion was identified using flanking telomeric and centromeric clones. We observed 28 cases out of 223 informative samples

(12.6%) contained deletions in any of the three loci examined. Interestingly,

25 deletion of FOXP3 locus was found in all of the 28 cases (Figure 2.8B and Table

2.2). These data suggest that FOXP3 is likely one of the most critical genes in

the Xp11.2 region studied. The FOXP3 protein was undetectable in 26/28 cases.

The two cases with both deletion and FOXP3 expression had X polysomy with 3

and 4 X-chromosomes respectively (Table 2.2).

Thirdly, we isolated DNA from normal and cancerous tissues from 50

patients with invasive ductal carcinoma and amplified all 11 coding and

- boundary regions by PCR. The PCR products were sequenced.

When the cancerous tissues were compared with normal tissues, about 38%

(19/50) of the samples showed somatic mutations (Figure 2.9). In addition to two

nonsense mutations, 15 mutations resulted in non-conservative replacement of

amino acids in domains that are either known to be critical for FOXP3 function,

as judged from the pattern of mutation in IPEX patients (Ziegler, 2006), or in the

conserved zinc finger domain that has so far not been implicated (Figure 2.9A).

Three cases had mutations in and two cases had exon mutations near

intron that may potentially affect RNA splicing. These data demonstrate that the

FOXP3 locus is targeted in breast cancer by a number of different mechanisms, including deletion, somatic mutation and other unknown mechanisms of

silencing. Since the samples used for mutational analysis were limited and did

not completely overlap with those used for FISH, it is unclear whether deletion

and mutation of the FOXP3 locus acts in concert to inactivate the FOXP3 locus.

As described in Chapter 3, we showed that FoxP3 is a transcriptional

repressor for HER2/ErbB2 gene. To directly test whether FOXP3 mutations affect

26 the function of FOXP3, we chose two representative somatic FOXP3 mutants isolated in cancer cells and test their repressor activity for HER2 promoter. One

mutation (338P>L) resides in the signature forkhead domain that is often mutated

in the IPEX patient, while the other double mutation (204C>R/205E>K) is from

zinc finger domain that has not been implicated in IPEX patients. As shown in

Figure 2.9C, both mutations significantly reduced the repressor activity of

FOXP3.

2.5 Discussion

Tumor suppressor genes (TSGs) are the most critical cancer related

genes (Macdonald et al., 2004; Vogelstein and Kinzler, 2002). Because they

usually control the normal rate of cell growth and turnover as well as maintain

genetic integrity, loss-of-function of TSGs is one of the key steps in

turmorigenesis. Two aberrant genetic alterations are required to silence a

particular TSG in cancer cells (Knudson, 2001; Knudson, 1971), as a result,

those individuals who carry a germ line mutation in one allele of a TSG have a

much higher risk to develop tumors in their life-time compared to a person with

the two normal alleles. Most putative TSGs are involved in human genetic

syndromes, such as BRCA1 and BRCA2 in familial breast and ovary cancers, RB

in familial retinoblastoma, and p53 in Li-Fraumeni syndrome. A very interesting

fact is that most TSGs were first identified by studying human hereditary cancers

(Vogelstein and Kinzler, 2002). Genetically engineered mouse (GEM) models

were subsequently established to analyze the effects of the defects of these

TSGs on experimental animals (Ghebranious and Donehower, 1998;

27 Hennighausen, 2000). However, we report here a reversed case in which we first

found the tumorigenic phenotype in a mutant mouse model and then showed the

relevant genetic evidence in human cancers.

Mouse cancer models with defective TSG

A tremendous amount of evidence shows that loss-of-function TSG is

critical for tumorigenesis in both human and mouse (Ghebranious and

Donehower, 1998; Hennighausen, 2000; Vogelstein and Kinzler, 2002), however the defective TSG usually gives different tumor spectrums in human and mouse, or even in different strains of in-bred mice (Donehower et al., 1995; Ludwig et al.,

2001a; Venkatachalam and Donehower, 1998). Germ line mutations of BRCA1 and BRCA2 cause breast and ovary cancers in human and mutations of p53 are found in 50% human breast cancers, but genetic knock-out of these genes in mice usually causes lymphomas and sarcomas (Ludwig et al., 2001a;

Venkatachalam and Donehower, 1998; Vogelstein and Kinzler, 2002). Ludwig et al. reported that among 92 tumors developed in 76 Brca1 hypomorphic mutant mice, 35% were lymphomas, 11% sarcomas, 13% breast cancer, 14% lung cancer, 14% liver cancer, 2% uterus cancers and only one case of benign ovary tumo (Ludwig et al., 2001a)r. When these hypomorphic Brca1 mutant mice were backcrossed to either p53-/- or p53+/- mice, no dramatic change in their tumor spectrum was observed. Cressman et al. had a similar finding by using Brca1+/- heterozygous mutant in a p53-/- background, in which only 10% of the mice developed mammary cancer compared to a 69% incidence of lymphoma. While no mammary cancers were detected in the Brca1+/- heterozygous mutant mice

28 in a p53+/- background strain (Cressman et al., 1999). Taking advantage of a tissue specific conditional gene targeting technique, Chu-xia Deng’s group first showed that more mammary cancers developed in the genetically engineered mouse model that is conditional knockout Brca1 in mammary epithelial cell

(Brodie et al., 2001; Xu et al., 1999). They reported that mammary tumors were detected in 25% of all mice by 2-years of age. And they reported a dramatic increased tumor incidence in a p53+/- background, which all of mice developed mammary tumors by 15-months old of age (Xu et al., 2001). This result is consistent with genetic evidence found in human breast cancer which show that many BRCA1 associated breast cancers also show a high incidence of p53 mutation (Vogelstein and Kinzler, 2002). Mammary tissue specific knockout

Brca2 mice showed a higher breast tumor incidence (77%) with a 1.4 year median tumor-free survival time (Ludwig et al., 2001b). More uniform mammary tumor histology, solid and nodular tumor pattern, was found in Brca2 conditional knockout mice compare with Brca1 mutant mice, which suggested that Brca2 targeted a more specific pathway in breast tumorigenesis. On the other hand, varying tumor spectrums were found in p53 mutant mouse strains with different genetic backgrounds (Donehower et al., 1995; Harvey et al., 1993). Lawrence A.

Donehower and his colleague first reported a p53 knockout mouse model in a mixed C57BL/6X129/Sv genetic background (Donehower et al., 1992), in which lymphoma and sarcomas counted for more than half of the tumors raised in both p53-/- and p53+/- mice. In a pure 129/Sv background, more teratocarcinomas were detected, but mammary tumors were rarely observed in the both strains

29 (Harvey et al., 1993). However, Joseph Jerry’s group (Kuperwasser et al., 2000)

backcrossed either p53-/- or p53+/- mice in the BALB/c background and found a

55% incidence of mammary cancer in the p53+/- BALB/c mice with a one-year median tumor-free survival time. Sarcomas and lymphomas were the most frequent tumors in the p53-/- mice. Very few mammary tumors were found in the p53-/- mice which grew other tumors and died by 26 weeks of age.

Interestingly, using the same BALB/c background we report here that

Foxp3sf/+ mice represent another defective TSG mouse model with a high

incidence of mammary cancer. We found that 60% of the tumors detected in

Foxp3sf/+ mice were mammary cancers, 18% were lymphomas, and 8% were

sarcomas. Compared to p53+/- mice, Foxp3sf/+ mice had a longer median tumor- free survival time (1.5 year v.s. 1 year). Second, no metastases were detected in

p53+/- BALB/c mice, while one-third of the mammary cancers metastasized to the lung in Foxp3sf/+ mice. Thirdly, the mammary tumor histology patterns in

Foxp3sf/+ mice were more diverse than those found in p53+/- mice. The

differences between these two mutants in the same BALB/c strain indicate that

Foxp3 and p53 may be involved in different pathways in mammary tumorigenesis.

X chromosome and breast cancer

X chromosome and X linked genes are very interesting topics in cancer

genetics, because of their relevance to female-specific cancers. Although there is

a very rare incidence of male breast cancers, an increased incidence of breast

cancer has been reported in Klinefelter’s syndrome (XXY genotype) patients

(Spatz et al., 2004). One of the key biological issues involved here is X

30 inactivation which compensates for the imbalance of the number of X

chromosome in female and male mammals. As most X linked genes are silenced

on Xi (inactivated X chromosome), any aberrant genetic changes in an X linked

TSG on Xa (activated X chromosome) would finally shut down this gene’s

expression in the cell. Based on this scenario, cells carrying a defective TSG on

Xa would subsequently possess the capability of unrestrained growth. As a result,

more cells would carry the Xa with a defective TSG than those with a normal

TSG allele. This nonrandom X inactivation is called skewed X inactivation (Spatz

et al., 2004). Some X linked recessive disorders in females have shown skewed

X inactivation. For instance, a female leukemia patient carrying heterozygous

WAS (Wiskott-Aldrich syndrome) gene displayed skewed X inactivation of the

normal allele in all malignant cells (Parolini et al., 1998). We also found that the

wild type Foxp3 allele was subject to skewed X inactivation in the mammary

cancers developed in Foxp3 mutant mice. Therefore an interesting question is whether the same biological effect exists in human breast cancers. It would be of great interest to study breast cancer incidence in family membranes of IPEX syndrome patients (, Polyendopathy, Enterophathy, X-

linked), in which germ-line mutations of FOXP3 have been reported (Bennett et

al., 2001; Wildin et al., 2001; Ziegler, 2006). Unfortunately we were not able to

access those patients. A Norwegian group (Kristiansen et al., 2005; Kristiansen

et al., 2002) reported skewed X inaction in young familial breast cancer patients

which suggests that unknown X linked suppressor genes are related to breast

cancer.

31 Other lines of evidence pointing out the involvement of X chromosome

abnormality in breast cancers includes a study connecting BRCA1 to X inactivation in which it was demonstrated that BRCA1 is involved in X inactivation through an interaction with XIST RNA (Ganesan et al., 2002).

Interestingly, duplication of Xa and the loss of Xi frequently occur in BRCA1- deficient cancers and basal-like breast cancers (BLC) - a group of high-grad aneuploid invasive ductal carcinomas characterized by expression of specific basal epithelial cytokeratins but usually without expression of ER, PR and HER2

(Richardson et al., 2006). Surprisingly, the increased expression of most X linked genes was not found in cancer cells with Xa duplication (Richardson et al.,

2006). On the contrary, duplication of Xi (inactivated X chromosome) was also found in 25% (5/20) non-BLCs (Richardson et al., 2006). Sirchia et al also reported the polysomy X in some breast cancer cell lines(Sirchia et al., 2005).

In our study of 223 breast cancers, we found that 40% of the cancers

carried more than two X chromosomes. Among them 21% were X trisomy, and

19% contained 4 to 6 X chromosomes. This abnormal duplication of X

chromosomes in breast cancers, called isodisomy, might be caused by genomic

instability in malignant cells. Interestingly, isodisomy could also play a specific

and critical role to accelerate breast tumorigenesis as isodisomy was not

detected in every chromosome but most frequently found in chromosome X, 14

and 17 (Richardson et al., 2006).

By examining a large panel of human sporadic cancers, we found that

most cancers lost FOXP3 expression, and detected a high incidence of somatic

32 mutations (38%, 19/50) and deletions (12.6%, 28/223) of FOXP3. Although at

this point we cannot definitively conclude that these two mechanisms act in

concert to inactivate the FOXP3 locus, the prevalent genetic alterations in the X

linked FOXP3 gene and the abnormal duplication of X chromosomes seen in

breast cancers could support the following hypothesis: (1) Aberrant genetic

changes at the FOXP3 locus may occur earlier than X chromosome duplication

during malignancy transformation (in our data, FOXP3 deletion almost evenly distributes between polysomy X, 15/28, and diploid X, 13/28). (2) Because

FOXP3 has been shown to be subject to X inactivation (Carrel and Willard, 2005;

Tommasini et al., 2002), and if the assumption above is justified, we could hypothesize that cancer cells are more likely to carry the mutant FOXP3 allele on

Xa (activated X chromosome), which could explain why loss of expression of

FOXP3 was found even in many cancer cells carrying duplicated Xas. This was

reflected to some extent in our data which showed that more somatic mutations

(38%) occurred compared with deletions (12.6%) of FOXP3. Nevertheless,

further work will be required to decipher this intriguing mechanism behind

silencing X linked TSGs.

33

Figure 2.1. Increased susceptibility to malignancies in heterozygous

Foxp3sf mutant mice.

A. Representative breast cancers developed in the Foxp3sf/+Otcspf/+ female mice.

The top panel shows the gross anatomy while the lower panel shows histology of

local and lung metastatic lesions of a breast cancer. B. Cancer-free survival analysis of Foxp3sf/+, Foxp3sf/+Otcspf/+, Otcspf/+ and WT littermates. Mice were

sacrificed when moribund to identify the tissue origins of cancers. log-rank test :

Foxp3sf/+ vs. wt, P<0.0001; Foxp3sf/+ vs. Otcspf/+ , P=0.0003; Foxp3sf/+ vs.

Foxp3sf/+Otcspf/+ , P=0 .9526; Foxp3sf/+Otcspf/+ vs. WT, P=0.0001; Foxp3sf/+Otcspf/+ vs. Otcspf/+, P=0.0001; Otcspf/+ vs. WT, P=0.4164. C. Increased susceptibility of

Foxp3sf/+ mice to carcinogen DMBA and progesterone as described (Aldaz et al.,

1996). The diagram on the top depicts experimental protocol, while survival

analysis was shown at the bottom panel. log-rank test: Foxp3sf/+ vs. WT,

P<0.0001; Foxp3sf/+Otcspf/+ vs. Otcspf/+, P=0.0005; Otcspf/+ vs. WT, P=0.8157. In

B& C, those mice that were observed for only part of the duration were

incorporated as censored samples and were marked with a cross in the Kaplan-

Meier survival curves.

34

Figure 2.1. Increased susceptibility to malignancies in heterozygous Foxp3sf mutant mice.

35

Malignancy Types Wild-type Foxp3sf/+ Otcspf/+ Foxp3sfOtcspf/++ Mammary carcinoma 5 7 2 19 Adenocarcinoma 3 5 2 12 Adenosquamous 2 2 0 7 carcinoma Multifocal cancer 0 1 0 7 lesions Lung metastasis 0 2 0 8 Lymphoma 2 1 0 7 Liver Tumor 0 0 0 4 Adenoma 0 0 0 3 Carcinoma 0 0 0 1 Sarcoma 0 0 0 3 Hemangiosarcoma 0 0 0 2 Leiomyosarcoma 0 0 0 1 Uterus adenocarcinoma 0 0 0 2 Multiple organs 0 0 0 12 tumorigenesis** *: The Arabic numeral in table indicates the number of mice carrying the corresponding tumorigenic characteristics. **: Mice bearing several tumors derived from different origins of organs.

Table 2.1 Histological features of spontaneous cancers in wild type and

Foxp3sf mutant BALB/c mice*

(Dr. Huiming Zhang and Histology Core Facility in Department of Pathology in Ohio State University accomplished all of histological studies shown in table 2.1, figure 2.1A, figure 2.2 and figure 2.3)

36 Figure 2.2. Histopathology of spontaneous tumors from heterozygous

Foxp3sf mice.

A and B. Representative lymphoma developed in Foxp3sf/+Otcspf/+ female mice.

Gross anatomy with enlarged spleen and are shown. The arrow in B

indicates the lymph node. C to F. Histological sections of a T cell lymphoma are

shown. Immunophenotype of the tumor is B220 negative and CD3 positive

shown in Insets of C. Infiltrated malignant lymphocytes to liver, lung and intestine

are shown in D, E and F respectively. G and H. H/E and reticuline histology

sections of one liver carcinoma are shown respectively. I. Uterus

adenocarcinoma invading the uterus wall. J and K. Leiomyosarcoma: tumor cells

forming a spindle pattern in J and sarcoma invading a mammary gland in K are

shown. L. Hemangiosarcoma: massive newly-formed blood vessels and bleeding

are shown. Original magnifications: C with insets 200X; D, E, F, I, L 100X; G, H,

J, K 400X.

37

Figure 2.2. Histopathology of spontaneous tumors from heterozygous

Foxp3sf mice.

38 Figure 2.3. Histopathology of mammary glands and tumors in mice.

A. The normal mammary gland is surrounded by fat tissue from a wild-type female mouse; inset shows the higher magnification of one ductule. B. A hyperplasic lesion with thickened stroma and abnormal ducts and alveoli from one Foxp3sf/+Otcspf/+ female mouse. Inset shows multiples layers of atypical epithelia cells in one duct. C. A sarcoma invading mammary gland with bleeding.

D to I, Three histological types of mammary carcinomas found in Foxp3 mutant female mice. The representative solid, papillary and squamous carcinomas are shown in D, E and F respectively. The square frame encircled areas in D, E and

F were enlarged with a higher magnification and are shown in G, H and I separately. J and K. Immunohistochemical study of estrogen and progesterone receptors in mammary tissues. J and inset show estrogen receptor positive cells in a mammary tumor and adjacent mammary gland, K shows a progesterone receptor positive tumor with a nearby gland at the lower-left corner of the figure.

Original magnifications: A, B, C, D, E, F 200X; insets in A and B, G, H, I, J, K

400X.

39

Figure 2.3. Histopathology of mammary glands and tumors in mice.

40

Figure 2.4 No Otc expression detected in mammary gland.

A. The procedure of laser capture microdissection (LCM). From right to left, three panels show one mammary gland before and after LCM. B. The realtime RT-

PCR profile shows there were no detectable Otc transcripts in mammary glands;

Otc transcripts are highly expressed in liver which was used as positive control.

C. Histochemical studies of the enzymatic activity of OTC in the liver from either an Otc mutant hemizygous mouse or wild-type mouse and the mammary gland from a wild-type female mouse. No OTC activity was found in mammary glands.

41

Figure 2.5 Foxp3 is silenced in mammary cancer by skewed X-inactivion

A. Defective Foxp3 expression in breast cancer. Total RNA extracted from cells procured by LCM was subjected to quantitative real-time RT-PCR using primers specific for Foxp3, Cd3, Hprt and CK19. In the left panel, fluorescence intensity

(∆Rn) was plotted versus cycle number. B. Mean and S.D. from three individual mice per group are presented in the right panel (P<0.0001, one-way ANOVA test when either internal standards were used). C. Specific silencing of the WT allele in breast cancer cells. Foxp3 transcripts were amplified from micro-dissected breast cancers or normal breast epithelium by two rounds of anchored PCR and were cloned into the Topo vector and sequenced. Left panels show the chromatograms of the mutant (top) and WT transcripts. The right panel shows the number of clones with sequences of WT or mutant alleles in the breast cancer and normal epithelium. A total of 20 clones were sequenced from each group. Data shown are from pooled samples that lack CD3 transcripts. n.d., not detectable. (P<0.001, Chi-square test)

42

Figure 2.5 Foxp3 is silenced in mammary cancer by skewed X-inactivion

43 Figure 2.6. Foxp3 inhibits the growth and tumorigenicity of multiple

breast cancer cell lines.

A. Breast cancer cell lines MCF-7, SKBr3 and TSA were transfected with equal

concentrations of either vector alone (Vector), Foxp3 or Otc cDNA. After 3

weeks of G-418 selection, the drug resistant clones were visualized by crystal

violet dye. B. Summary of the colony numbers in three independent

experiments as described in A. Data shown are means and S.D. (Vector and Otc

vs. FoxP3: P<0.001, T test) C. Expression of Foxp3 reduces growth rate of

tumors. Syngeneic BALB/c mice were injected with 5x105/mouse Foxp3 or

vector-transfected TSA cells in the flank and the sizes of the local tumor mass

were measured using a caliper. Data shown are means and S.D. and have been

repeated once. (P<0.0001, one-way ANOVA test) D. The survival of tumor-

bearing mice was monitored over a 7-week period. As in C, except that 106 of tumor cells were injected per mouse. Mice were euthanized when they became moribund. A significant difference was observed in the survival of mice bearing the different types of tumor cells (P=0.0015, log-rank test).

44

Figure 2.6. Foxp3 inhibits the growth and tumorigenicity of multiple breast cancer cell lines.

45 Figure 2.7 Human breast cancers have lost FOXP3 expression.

A. Down-regulation of FOXP3 protein among human breast cancer cells.

Photographs in the top and middle panels show immunohistochemical staining of

normal and carcinoma tissues from the same patient. The numbers of FOXP3

positive tissues are shown in the lower panel. The samples with nuclear staining

by the anti-Foxp3 antibody were scored as positive. P value derived from Chi-

square test. B. Comparison of FOXP3 between PBL (peripheral

blood) samples, HMEC (human mammary epithelial cells) and 9 human breast

cancer cell lines listed in figure. Top panel shows the RT-PCR amplified full-

length FOXP3 mRNA isoforms, while the lower panel shows the quantitative

difference, as revealed by real-time PCR, P<0.0001 T test.

46

Figure 2.7 Human breast cancers have lost FOXP3 expression.

47

Figure 2.8 Deletion of FOXP3 locus in human breast cancer cells.

Breast cancer tissue microarray samples were analyzed by using FISH probes surrounding a 10MB region in Xp11.2. Details of the method applied is described in section 2.3. A. The genomic structure of X chromosome is illustrated in the left panel. 3 probes are indicated as different color bars and their positions are labeled in the figure. The deletions in 28 samples are summarized as the 28 corresponding color bars in the right panel. As shown in the figure,

FOXP3 locus (yellow) is a common deletion region in all 28 cases. B. FISH

(fluorescence in-situ hybridization) picture of a representative breast cancer with deletion of FOXP3 locus shows only one FOXP3 (orange) probe signal with two

X chromosome signals indicated by CEPX (green).

Figure 2.8 and table 2.2 are presented here through the courtesy of Dr.

Carl Morrison in Department of Pathology in Ohio State University.

48

Figure 2.8 Deletion of FOXP3 locus in human breast cancer cells.

49 Case FOXP3 HER2 CEPX b RP11-344O14 c X Chromosome status number IHC a IHC d 1 - - 2 81 Diploid X 2 - - 2 92 Diploid X 3 - ++ 2 77 Diploid X 4 - + 2 84 Diploid X 5 - ++ 2 47 Diploid X 6 - ++ 3 82 Polysomy X 7 - - 3 74 Polysomy X 8 - - 4 56 Polysomy X 9 - + 4 61 Polysomy X 10 - +++ 4 64 Polysomy X 11 - +++ 4 52 Polysomy X 12 - - 2 89 Diploid X 13 - + 4 61 Polysomy X 14 - - 3 72 Polysomy X 15 - + 2 95 Diploid X 16 - ++ 4 81 Polysomy X 17 - + 2 93 Diploid X 18 + - 4 80 Polysomy X 19 - + 2 81 Diploid X 20 - +++ 2 74 Diploid X 21 - - 2 78 Diploid X 22 - + 4 57 Polysomy X 23 - +++ 4 66 Polysomy X 24 - + 2 90 Diploid X 25 - - 2 93 Diploid X 26 - + 3 57 Polysomy X 27 - +++ 3 46 Polysomy X 28 + - 3 68 Polysomy X *: The 28 cases listed in this table are same FOXP3 locus deletion cases demonstrated in Figure 2.8. (a): Immunohistochemistry for nuclear FOXP3 staining. (b): X chromosome centromere copy number. (c): % cells in cancerous section with reduced copy number of FOXP3 on X chromosome. (d): HER2 protein level expressed in cancer tissue is measured by immunohistochemical stain and scored by following previous description (Yaziji et al., 2004);

Table 2.2 FOXP3 deletions in 28 cases of breast cancer

50 Figure 2.9 Somatic mutations of FOXP3 in the breast cancer samples.

Comparison of DNA sequences from normal and cancerous tissues of 50 breast cancer patients identified somatic mutations. A. Schematic cartoon shows all of somatic mutations of FOXP3 found in 50 cases. Arrows in black indicate mutations in encoding regions, while ones in blue are mutations in introns. B. The representative chromatogram of DNA sequencing shows a mutation in codon 87, causing a change of glycine to aspartic acid. C. FOXP3 mutations reduced its activity in repressing HER2/ErbB2 promoter. As we show in Chapter 3, one of the critical functions of FOXP3 in breast tumorigenesis is to transcriptionally repress HER2 oncogene expression in breast epithelial cells.

The 2.1kb human HER2 promoter linked to luciferase reporter (0.25 µg/well) was co-transfected with 1µg/well of either vector control, WT or two mutants of the

FOXP3 isolated from breast cancer tissues (338 P>L in the FKH domain, or

204C>R/205E>K in the zinc finger domain). Luciferase activity was measured 48 hours after transfection. Data shown are representative of at least 3 independent experiments. The difference between WT and 338 P> L (P=0.0004) and that between WT and 204C>R/205E>K (P=0.003) are highly significant.

Dr. Lizhong Wang kindly helped me to finish all of FOXP3 sequencing work that is shown in figure 2.9 and table 2.3.

51

les. p FOXP3 in breast cancer sam ure 2.9 Somatic mutations of g Fi

52

Table 2.3. Summary of somatic FOXP3 mutations in breast cancer

Comparison of FOXP3 sequences of cancerous tissues to that of normal tissues from various origins, including tumor-free lymph nodes, skin, glands and occasionally fibroblasts identified somatic mutations. Briefly, genomic DNA were isolated from cancerous and normal tissues from 50 breast cancer patients and amplified with primers for individual exons and intron-exon boundary regions. All

DNA were isolated from formalin-fixed tissues and were amplified by PCR.

Comparing sequences from normal and cancerous samples identified somatic mutations. The overwhelming majority of data are from bulk sequencing of PCR products. In some experiments, the PCR products were cloned and the mutations were verified by sequencing multiple clones. The mutations were scored only if found in more than one clone.

53

mutations in breast cancer 3 FOXP somatic f o y Table 2.3. Summar

54

mutations in breast cancer 3 FOXP somatic f o y Table 2.3. Summar

55

CHAPTER 3

FOXP3 SUPPRESSES BREAST CANCER BY REPRESSING HER2/ERBB2 ONCOGENE

3.1 Abstract

HER2/ErbB2 is one of most critical oncogenes identified in breast cancer

(Slamon et al., 1987; van de Vijver et al., 1988). The dominant murine ErbB2 transgenic mouse models develop aggressive mammary tumors with high penetrance (Guy et al., 1992; Muller et al., 1988), however no spontaneous mouse mammary cancers have been reported with overexpression ErbB2.

Although HER2 gene amplification is found in 25 % of human breast cancers

(Kallioniemi et al., 1992; Slamon et al., 1987), increased activity of the

HER2/ErbB2 promoter has also been shown to have an important role in the

over-expression of this oncogene in breast cancer (Hurst, 2001; Scott et al.,

2000; Xing et al., 2000). Here our data show that HER2/ErbB2 oncogene is overexpressed in spontaneous mammary cancers in Foxp3sf/+ mutant mice.

FoxP3 specifically binds upstream of HER2/ErbB2 and transcriptionally

represses HER2/ErbB2 expressions in cancer cells. Silencing of FOXP3 by

siRNA in normal human mammary epithelial cells (HMEC) increases

HER2/ErbB2 expression. Furthermore, in both amplified and non-amplified HER2

56 breast cancers, FOXP3 expression status is inversely correlated with HER2

expression levels, which indicates that a genetic lesion at FOXP3 locus, at least

in part, causes HER2/ErbB2 overexpression in human breast cancers. Taken

together, our data indicate that FoxP3 suppresses breast cancer by repressing

HER2/ErbB2 oncogene expression.

3.2 Introduction

Proto-oncogene HER2/ErbB2 belongs to the epidermal growth factor

receptor tyrosine kinase family (Yarden and Sliwkowski, 2001). Terminologically,

ErbB2 and Neu denote the cellular proto-oncogene in mouse and rat respectively, while HER2 is often used for the orthologue of the human proto-oncogene. Since both murine and human subjects were involved in this study, we refer to this gene as HER2/ErbB2 in the text. HER2/ErbB2 is over expressed and amplified in

25 to 30% of human breast cancers (Slamon et al., 1987; van de Vijver et al.,

1988). Herceptin, a recombinant humanized monoclonal anti-HER2/ErbB2 antibody, has been shown to have therapeutic efficacy in most breast cancer patients with overexpression of HER2/ErbB2 (Slamon et al., 2001). Gene amplification of HER2/ErbB2 is tightly correlated with overexpression of this oncogene in breast cancers (Kallioniemi et al., 1992). However, several lines of evidence have shown that HER2/ErbB2 overexpression is found in many breast cancers without HER2/ErbB2 gene amplification (Bofin et al., 2004; Jimenez et al., 2000) and that HER2/ErbB2 mRNA levels are increased per gene copy in the gene-amplified cancer cells (Kraus et al., 1987). Interestingly, several viral onco-

57 products such as those encoded by the adenovius type 5E1A gene and the simian virus 40 large T antigen have been shown to repress HER2/ErbB2 promoter activity, although the mechanism is unclear (Hung et al., 1995). More recently, Ets transcriptional factor PEA3 has been shown to repress HER2/ErbB2 expression (Xing et al., 2000), though genetic evidence of PEA3 in human breast cancers is not available yet.

In this report, we demonstrate that HER2/ErbB2 is over expressed in mammary cancers developed in mutant Foxp3 heterozygous mice. We also show that FoxP3 binds upstream of HER2/ErbB2 gene at forkhead domain DNA binding motifs (5’-RYMAAYA-3’; R=A,G, Y=C,T, M=A,C) in regions conserved between human and mouse. We find that FoxP3 represses expression of

HER2/ErbB2 in breast cancer cell lines, and silencing FOXP3 by siRNA in normal human epithelial cells causes increased expression of HER2/ErbB2 transcripts. An inverse association between HER2/ErbB2 and FOXP3 expression was found in both amplified and non-amplified HER2/ErbB2 breast cancers, which indicates that loss of expression of FOXP3 could be one of the genetic lesions causing overexpression of HER2/ErbB2 in cancers. Thus FOXP3 might prove to be a potential prognostic marker for human breast cancer.

58

3.3 Materials and methods

Cells, antibodies and animals

Described in Chapter 2.

Laser capture microdissection, mRNA extraction and quantitative real-time

PCR

Described in Chapter 2.

Luciferase Reporter assay

The promoter region of ErbB2 was amplified by using a reverse primer with

BlgII linker at the 5’ end: GGGAGATCTCAATCTCAGCTCCACAACTTCAC; and three forward primers with a SacI linker at the 5’ end:

0.8kb upstream: GGGGAGCTCTGAGAACTGGGTAAAGTCAGA,

1.2kb upstream: GGGGAGCTCGAGGGAAGATACGAACTCAGGTC, and 1.8kb upstream: GGGGAGCTCTTTGTCACATGTATGTGTTGAAC. The

PCR products were cloned into the pGL-2-Basic vector (Promega) at SacI/BlgII sites. The human HER2 promoter region (0.5kb) was cloned following a previously reported method (Xing et al., 2000). The 2.1kb promoter region was cloned into pGL2 at XhoI/MluI sites by using the following primers: XhoI

Reverse:TTTCTCGAGAATGGAGGGGAATCTCAGCTT, and

MluI forward:TTTACGCGTCCATACTCGGCCAACTTTGT. All inserts were confirmed by sequencing to match the following Genebank sequences: ErbB2

(Mus musculus chromosome 11 genomic contig, NT_039521) and HER2 transcript variant1 (Source Sequence BM678576, CN409735,M11730,X03363).

59 For the luciferase assay, 1 x 104cells/well of cells were seeded in 24-well-plates.

0.3 µg of luciferase construct and 0.02 µg of pRL-TK (Promega) were transiently co-transfected with 1µg of Foxp3 expression plasmid or the empty vector respectively. After incubation for 48 hours, the cells were harvested with Passive

Lysis Buffer (Promega), and luciferase activities of cell extracts were measured with the use of the Dual luciferase assay system (Promega). The specific amounts of DNA used in these experiments are described in the text and legend.

Chromatin immunoprecipitation (ChIP)

ChIP was carried out according to a published procedure (Im et al., 2004).

Briefly, the Foxp3-V5-transfected TSA cells were sonicated and fixed with 1% paraformaldehyde. The anti-V5 antibodies or control mouse IgG were used to pulled down chromatin associated with Foxp3-V5. The amounts of specific DNA fragments were quantitated by real-time PCR and normalized against the genomic DNA preparation from the same cells. The PCR primers are listed in

Table 3.4.

FOXP3 silencing lentiviral vector

The lentivirus-based siRNA expressing vectors were created by introducing the murine U6 RNA polymerase III promoter and a murine phosphoglycerate kinase promoter (pGK)-driven EGFP expression cassette into a vector with a pLenti6/V5-D-TOPO back bone without a CMV promoter. A hairpin siRNA sequence of FOXP3 (target sequence at the region of 1256 to

1274 nucleotides; 5’-GCAGCGGACACTCAATGAG-3’) was cloned into the lentiviral siRNA expressing vectors by restriction sites of ApaI and EcoRI. 60

Immunohistochemistry and Fluorescence In-situ Hybridization (FISH)

HER2 expression was performed using PathwayTM HER2 (Clone CB11)

(Ventana Medical Systems, Inc., Tucson, AZ) on the BenchMark® XT automated system per the manufacturer’s recommended protocol. FISH for HER-2 amplification was done using the PathVysion HER2 DNA Probe kit (Vysis, Dover,

Ill.) in accordance with the manufacturer’s guidelines. All FISH were done using standard protocols optimized for breast cancer specimens. Briefly, formalin fixed, paraffin-embedded tissue microarray blocks were cut into 3 to 4 µm thick sections, incubated over night at 560C, deparaffinized, washed, digested with

protease, formalin fixed, denatured, and hybridized at 37°C for 16 hours. The

slides were then washed in a post-hybridization wash, counter stained with 4'-6-

Diamidino-2-phenylindole (DAPI), and covered with a coverslip. Specimens were

evaluated with an Olympus BX51 microscope (Olympus Optical Company, LTD.,

Japan) under oil immersion at x150 magnification using the recommended filters.

3.4 Results

3.4.1. Mammary tumors in Foxp3sf/+ mutant mice overexpress HER2/ErbB2

oncogene.

HER2/ErbB2 over-expression is critical for the growth of multiple breast

cancer cell lines and confers a poor prognosis in breast cancer patients

(Kallioniemi et al., 1992; Slamon et al., 1987). To understand the mechanism by

which disruption of Foxp3 promoted tumorigenesis, we analyzed expression of

61 HER2/ErbB2 in normal and cancerous breast tissues. As shown in Figure 3.1A

and Table 3.1, mammary cancer tissue from the Foxp3sf/+Otcspf/+ mice showed

significant over-expression of HER2/ErbB2 as revealed by

immunohistochemistry. Using real-time RT-PCR, 8 to12 folds more HER2/ErbB2

mRNA was found in the cancer cells compared to normal epithelium (Figure

3.1A). There was also more HER2/ErbB2 mRNA in Foxp3sf/+Otcspf/+ epithelium than that found in WT female mice (Figure 3.1B), which indicates a potential gene dosage effect of Foxp3 on the regulation of HER2/ErbB2 expression in vivo.

3.4.2. FoxP3 is a transcriptional repressor of HER2/ErbB2 oncogene.

Increased activity of the HER2/ErbB2 promoter plays an important role in

the over-expression of this oncogene in breast cancer (Hurst, 2001; Scott et al.,

2000; Xing et al., 2000). One potential mechanism by which FoxP3 represses

HER2/ErbB2 is to inhibit the promoter activity. As shown in figure 3.2A, analysis

of the 5’ sequence of the HER2/ErbB2 gene revealed multiple binding motifs for

the forkhead domain, most of them clustered around 1.0-1.6 Kb upstream of the

transcription starting site (TSS). To test whether FoxP3 interacts with the

HER2/ErbB2 promoter, we used an anti-V5 antibody to precipitate sonicated

chromatin from TSA cells transfected with Foxp3-V5 cDNA and used real-time

PCR to quantitate the amounts of the specific HER2/ErbB2 promoter region

precipitated by the anti-V5 antibodies in comparison to those that bound to

mouse IgG control. As shown in Figure 3.2B, the anti-V5 antibodies pulled down

significantly higher amounts of mouse HER2/ErbB2 promoter DNA than the IgG

62 control. Moreover, the abundance of DNA correlated with the number of the

forkhead binding motifs, with the highest amount of precipitates near the 1.0-1.6

Kb region upstream of the TSS and the lowest amount of precipitates near the

8.5 Kb down-stream of TSS where no forkhead binding motifs were identified.

To test whether the binding correlated with transcriptional suppression mediated by FoxP3, we generated a luciferase reporter using the regions 1.8, 1.2 and 0.8

Kb upstream of the mouse ErbB2 TSS, and tested the ability of FoxP3 to repress

HER2/ErbB2 promoter activity. In three separate cell lines, we observed strong correlations between the number of forkhead DNA binding motifs and the repressor activity of FoxP3. Furthermore, we deleted two FoxP3 binding sites conserved between mouse and human (Figure 3.3A) by site-directed mutagenesis and measured its effect on FoxP3-mediated repression. As shown in figure 3.3B, deletion of either binding site substantially increased the

HER2/ErbB2 promoter activity in the presence of FoxP3 and thus alleviated

FoxP3-mediated repression. These results demonstrate that FoxP3 represses

HER2/ErbB2 by binding directly to the HER2/ErbB2 promoter region.

3.4.3 FOXP3 is an essential repressor of ErbB2/HER-2 in mammary

epithelia cells.

To test if HER2/ErbB2 protein levels could be repressed by FoxP3 in breast

cancer cells, we transfected Foxp3 cDNA into the mouse mammary cancer cell

line TSA and showed that both transcripts and protein levels of HER2/ErbB2

were reduced by Foxp3 (Figure 3.4B). To further prove whether FoxP3 is

essential to repress Her2/ErbB2 expression in breast epithelial cells, we silenced

63 FOXP3 gene in primary HMEC using a lentiviral vector expressing FOXP3

siRNA. As shown in Figure 3.4A, the FOXP3 siRNA reduced FOXP3 expression

by 100-fold while increasing HER2/ErbB2 mRNA by 7-fold. These results

demonstrate that FOXP3 is a repressor of HER2/ErbB2 in breast epithelial cells.

3.4.4 FOXP3 suppresses HER2/ErbB2 in human breast cancer.

We have obtained four lines of evidence that support a critical role for

FOXP3 in repression of HER2/ErbB2 oncogene in human breast cancer. First,

we compared the expression of FOXP3 with the scores of HER2/ErbB2 in breast

cancer tissues. As shown in Figure 3.5, down-regulation of FOXP3 was strongly

associated with over-expression of HER2/ErbB2, which supports a role for

FOXP3 defects in human HER2/ErbB2 over-expression in breast cancer.

Nevertheless, since many of the FOXP3 negative cells remained HER2/ErbB2

negative, it is likely that deregulation of FOXP3 is insufficient for HER2/ErbB2 up-

regulation. On the other hand, since only 3/82 FOXP3 positive cancer cases

expressed high levels of HER2/ErbB2, FOXP3 inactivation is likely essential for

HER2/ErbB2 up-regulation under most circumstances.

Second, since a major mechanism for HER2/ErbB2 up-regulation in breast

cancer is gene amplification, an intriguing issue is whether FOXP3 is capable of

repressing HER2/ErbB2 in cancer cells with an amplified HER2/ErbB2 gene. We produced a Tet-off line of BT474, a breast cancer cell line known to have

HER2/ErbB2 gene amplification (Kallioniemi et al., 1992), and transiently super- transfected it with a pBI-EGFP-FOXP3 vector (Figure 3.6B). After drug selection,

the cells were cultured either in the presence or absence of doxycycline. While

64 the cells cultured with doxycycline did not express FOXP3 (data not shown),

removal of doxycycline resulted in induction of FOXP3 in a significant fraction of

the cancer cells, which allowed us to compare HER2/ErbB2 levels in the FOXP3+ and FOXP3- cells in the same culture by flow cytometry. As shown in figure 3.6B,

FOXP3- cells had about a 10-fold higher level of HER2/ErbB2 protein on the cell surface in comparison to the FOXP3+ cells. These results demonstrate that

FOXP3 can suppress HER2/ErbB2 expression even in cells with multiple copies

of this oncogene.

Thirdly, we divided breast cancer samples based on their HER2/ErbB2

gene copy numbers and compared the FOXP3+ and FOXP3- cancer samples for

the relative amounts of cell surface HER2 expression. As shown in Figure 3.6A,

in each of the gene dose categories, FOXP3+ samples had reduced HER2 scores in comparison to the FOXP3- samples. These results strongly suggest a

critical role for FOXP3 in repressing HER2/ErbB2 expression even in the cases

with HER2/ErbB2 gene amplification.

Fourth, among the 232 samples that we screened for Xp11.2 deletions,

those with deletions encompassing FOXP3 locus had significantly higher HER2

scores compared to those without deletions (P=0.03) (Table 3.2). Likewise, we

compared the relative HER2 scores among the 50 samples in which we had

sequenced all FOXP3 encoding regions. As shown in Table 3.3, the mutations in the FOXP3 gene correlated with higher levels of HER2 (P=0.01). Also as shown in Chapter 2 figure 2.9C, FOXP3 mutations abrogated its repressive activity for

HER2/ErbB2 gene.

65 Taken together, we have demonstrated that FOXP3 is the first X-linked

breast cancer suppressor that represses HER2/ErbB2 oncogene. However, since

FOXP3 represses the growth of MCF-7 that does not over-express HER2/ErbB2,

additional FOXP3 targets may be involved in FOXP3-mediated cancer

suppression. Given the significant role of HER2/ErbB2 in pathogenesis of human breast cancer and the wide-spread defects of the FOXP3 locus, it is likely that

FOXP3 is an important suppressor for human breast cancer.

3.5 Discussion

HER2/ErbB2 is a major breast cancer oncogene and therapeutic target for

human breast cancer (Slamon et al., 1987; Slamon et al., 2001; van de Vijver et

al., 1988). HER2/ErbB2 over-expression indicates poor prognosis (Slamon et al.,

1987). The molecular lesions leading to the over-expression of HER2/ErbB2,

however, remain incompletely understood. Although the major contributing factor

of over-expression of HER2/ErbB2 is the amplification of this oncogene, it has

been shown that the transcripts of HER2/ErbB2 at per gene copy level are

increased in both breast cancer cells with and without amplified HER2/ErbB2

(Kraus et al., 1987). In addition, not all cancer cells with HER2 over-expression

show gene amplification (Bofin et al., 2004; Jimenez et al., 2000). An intriguing

possibility is defective expression of potential HER2/ErbB2 repressors. Recently,

PEA3 was reported by Xing et al. to be capable of binding to HER2/ErbB2

promoter region and repressing its expression (Xing et al., 2000). However, it is

less clear if defects in PEA3 are responsible for human HER2 over-expression in

66 breast cancers. Here we showed that Foxp3 scurfy mutation in mice resulted in

over-expression of ErbB2, the murine homologue of HER2. In addition,

transfection of Foxp3 repressed ErbB2 transcription. More importantly, chromatin

immunoprecipitation analysis revealed that Foxp3 bind directly to forkhead

domain DNA binding motifs (5’-RYMAAYA-3’; R=A,G, Y=C,T, M=A,C) in the 5’

region of the ErbB2 gene. By alignment of HER2/ErbB2 upstream regions in both

human and mouse, we found two-forkhead binding sites in the regions that are

conserved between human and mouse. Site-directed mutagenesis of theses two

binding sites ablated the repression of HER2/ErbB2 promoter activity by FOXP3,

which indicates that the binding of FOXP3 to HER2/ErbB2 promoter is specific,

and that other forkhead binding motifs in this region are very possibly non-

specific for FoxP3 function. Interestingly, we also found around a 2-fold decrease

in promoter activity of these mutated HER2/ErbB2 promoters compared to wild

type. Considering these regions share a high homology between human and

mouse, one can not exclude the possibility that these specific sequences are

important regulatory elements for HER2/ErbB2 expression, and that other critical transcriptional regulators may also interact with these two regions.

The inverse correlation between FOXP3 and HER2/ErbB2 levels found in

human breast cancer provides strong support for the transcriptional repression of

HER2/ErbB2 by FOXP3. Indeed, siRNA knock-down of FOXP3 in normal human mammary epithelial cells increased the expression of HER2/ErbB2 mRNA level in cells. Moreover, disruption of FOXP3 by either somatic mutations or deletions also correlated with high HER2 levels in human breast cancers. Combined with

67 the observation that mutant FOXP3 lost its ability to repress HER2 promoter activity, our data compellingly indicate that genetic lesions of the FOXP3 locus play a critical role in the overexpression of HER2/ErbB2 oncogene in breast cancers.

Taken together, we show that FoxP3 is a novel transcriptional repressor for HER2/ErbB2 oncogene and our data offers important insights for understanding the molecular pathogenesis of breast cancer.

68

Spontaneous Tumor Carcinogenesis induced Tumor**

Wild-type (n=5) Foxp3sf/+ (n=7) Wild-type (n=10) Foxp3sf/+ (n=12)

ErbB2

(-) 4 0 3 0

(+) 1 2 4 2

(++) 0 2 2 4

(+++) 0 3 1 5

*: The arabic numeral in table indicates the number of mice carrying mammary tumors. **: DMBA was used to induce mammary tumorigenesis in mice, the detail protocol described in Chapter 2.4.1. Statistic analysis: Chi-square test: P=0.0024

Table 3.1 ErbB2 Expression in Mouse Mammary Tumors*

69

Figure 3.1 Mammary tumors in Foxp3sf/+ mutant mice overexpress

HER2/ErbB2 oncogene.

A. Over-expression of ErbB2 in the mouse mammary cancers. The figure shows

immunohistochemical staining of a non-cancerous mammary gland and an

adjacent adenocarcinoma from a Foxp3sf/+Otcspf/+ mouse using anti-ErbB2

antibody. B. Relative expression levels of ErbB2 in normal mammary epithelium

of WT and Foxp3sf/+Otcspf/+ mice and the cancer tissues in the Foxp3sf/+Otcspf/+ mice, as revealed by real-time RT-PCR of LCM samples. Data shown are means and S.D. The expression of ErbB2 was normalized against either Hprt or CK19 genes. Highly significant differences were observed between cancerous and normal tissue (P<0.001, ANOVA test when either internal standards were used).

70

Figure 3.1 Mammary tumors in Foxp3sf/+ mutant mice overexpress

HER2/ErbB2 oncogene.

71 Figure 3.2 FoxP3 is a transcription repressor of HER2/ErbB2 oncogene.

A. The schematic diagram shows the 5’ region of the ErbB2 gene, including the

promoter, and exon 1 to exon 3. The forkhead binding motifs are illustrated with

small green bars, while the regions that were assayed by real-time PCR are marked in red bars. B. The amount of DNA precipitated by either control IgG or anti-V5 mAb, as expressed in relative units indicate the amounts of pull-down

DNA relative to a standard curve generated from a titration of total input DNA in the chromatin immuno-precipitation experiment (Im et al., 2004). C. Foxp3- mediated repression of the ErbB2 promotor requires forkhead binding motifs, as evaluated by dual-luciferaseR reporter assay. Three different cell lines were

transfected with either vector control or Foxp3 (1 µg/well) in conjunction with the

luciferase reporter driven by different 5’ promoter regions of the ErbB2 gene (0.6

µg/well). The promoter regions differed in the number of forkhead binding motifs, as illustrated in the diagram in the left. Forty-eight hours later, the cell lysates were harvested and measured for luciferase activity. Data shown are means and

S.D. of triplicates.

72

Figure 3.2 FoxP3 is a transcription repressor of HER2/ErbB2 oncogene.

73 Figure 3.3 Site-directed mutagenesis of Forkhead binding sites abrogates

repression of HER2/ErbB2 promoter by FoxP3.

A. The homologous sequences of two clusters of forkhead binding sites for

mouse and human HER2/ErbB2 promoter region are shown in the upper panel.

Sequence alignment was performed by using the NCBI BLAST program. The

deleted regions are underlined in the figure (deleted DNA sequence, mut A:

AAATCTGGGATCATTTA; Mut B: TTTGAATTTCAGATAAA). Diagrams of the

human HER2 and mouse ErbB2 promoters are shown in the lower panel. Arrows

indicate the homologous regions between the two promoters, and the two binding

sites were deleted individually. B. Mutations of either forkhead binding site

abrogates FOXP3-mediated suppression. The promoter activity was measured and normalized as detailed in Figure 3.2, except that the amount of promoter

DNA was 0.4 µg/sample.

74

Figure 3.3 Site-directed mutagenesis of Forkhead binding sites abrogates repression of HER2/ErbB2 promoter by FoxP3.

75 Figure 3.4 FOXP3 is an essential repressor of HER2/ErbB2 in mammary epithelia cells.

A. Silencing of FOXP3 resulted in up-regulation of HER2 in primary human mammary epithelial cells (HMEC). HMEC was transduced with lentiviral vector for either control sequence or FOXP3 siRNA. The untransfected cells were removed by selection with blasticidin. At one week after transduction, the levels of FOXP3 and HER2 transcripts were quantitated by real-time PCR. Data shown are means and SEM of relative levels of transcript (with that in the vector- transduced cells defined as 1.0) and are representative of three independent experiments. B. Transfection of Foxp3-V5 into TSA cells repressed expression of the ErbB2 locus. The left panel shows mRNA levels as measured by real-time

PCR. Data shown are means and S.D. of triplicates. The right panel shows the protein levels, as measured by Western blot of the cell lysates using anti-ErbB2 antibody. Actin was used as loading control. The amount of transfected Foxp3-

V5 was measured by Western blot using anti-V5 antibody.

The lentiviral siRNA for silencing FOXP3 was made and provided by Dr. Yan Liu.

76

Figure 3.4 FOXP3 is an essential repressor of HER2/ErbB2 in mammary epithelia cells.

77

Figure 3.5 An inverse correlation between FOXP3 and HER2/ErbB2 among human breast cancer samples.

Tissue microassay samples were stained with either anti-FOXP3 antibodies or anti-HER-2 antibodies and were scored by two pathologists in a double blind fashion. Three representative cases and summary data from 662 independent cases are presented. The P values of the chi-square tests are listed.

78 Figure 3.6 FOXP3 suppresses HER2/ErbB2 in human breast cancer.

A. Inverse correlations between FOXP3 expression and HER-2 scores in cells

with or without HER2 amplification. FISH determined the HER2 gene copy

number, while nuclear expression of FOXP3 was determined by immunohistochemistry. Data shown are means and SEM of HER2 scores of 425

cases of breast cancers grouped by HER2 copy number. Cell surface HER2

levels were scored using clinically accepted criteria. P values were generated by

Mann-Whitney test. B. In the Tet-off inducible FOXP3-expressing BT474, FOXP3

repressed HER2. BT474 cells were first transfected with pTet-Off vector (BD

Biosciences Clontech: Cat# K1620-A). G418 resistant clones with low

background and high doxycycline dependent induction were selected. BT474

Tet-off stable cells were transfected with pBI-EGFP-FOXP3, which was modified

from pBI-EGFP (BD Biosciences Clontech, Cat#6154-1), in conjunction with

pUB6 vector (Invitrogen, Cat# V250-2) that conveys blasticidin resistance. The

transfectants were selected by both blasticidin and G418 in doxycycline-

containing medium. The drug-resistant cells were cultured in the absence of

doxycycline for 5 days to induce FOXP3. The cells were stained for FOXP3 and

HER-2 proteins by flow cytometry. Data shown are histograms depicting HER2

levels among the gated FOXP3hi and FOXP3- cells and are representative of two independent experiments.

79

Figure 3.6 FOXP3 suppresses HER2/ErbB2 in human breast cancer.

80

HER2 Deletion (n=28) Non Deletion (n=195)

(-) 10 125

(+) 9 30

(++) 4 15

(+++) 5 25

*: The arabic numeral in table indicates the number of breast cancers in each category. Statistic analysis: Chi-square test P=0.03

Table 3.2 Effects of FOXP3 Deletions on HER2 Expression in Breast

Cancers*

HER2 Mutation (n=17) Non mutation (n=31)

(-) 2 19

(+) 6 6

(++) 5 3

(+++) 4 3

*: The arabic numeral in table indicates the number of breast cancers in each category. Statistic analysis: Chi-square test P=0.01

Table 3.3 Effects of FOXP3 Mutations on HER2 Expression in Breast

Cancers*

81

Position in promoter Primer sequence 5 ’to 3’ *

-3.2Kb Forward ACAGGCCACTGGTTTCAGAC Reverse TGAGGGAACTTCGAAGACAGA -2.2Kb Forward GGAGAAGGGACACCTTTGATCT Reverse GGGAATATCTGAGCCCTAGCAA -1.6Kb Forward AGCCCTCTTGTTCTACTTCTGG Reverse GACACTCTAGAAGCACTCAGCA -1.0Kb Forward CGGGCAATTCATCCTGGTAAAC Reverse GATATCACTCCTGAAGCCTGGT -0.4Kb Forward GAGAGTCTTGGAAGTCACCAGT Reverse GCAGTTCTCACCCACTTCCTAA +0.5Kb Forward GGGAACTCCTTGGGAAAGTTCT Reverse ACTGGAAGAGCTCTGAGAAAGC +1.1Kb Forward CGTGTTAGGCAAGCCCTCTA Reverse GGAATCCCAAAGCACACAGT +1.8Kb Forward TGTTGCCAAACAGCAGTCTC Reverse TCCATCCTGAAGAAGGCAAG +2.8Kb Forward TTGTGCTCTCTCTCTGCACTGT Reverse AGTCCGTTCCTGTTTGACAACT Exon 3 Forward ACATCCAGGAAGTCCAGGGATAC Reverse GCGGTGGTGACGTTGTCCAAA GAPDH Forward CCACCATCCGGGTTCCTATAAA Reverse TTGCACACTTCGCACCAGCAT *: all primer designs are based on DNA sequences downloaded from the NCBI data bank; the accession number for the mouse ErbB2 promoter region is NT_039521

Table 3.4 HER2/ErbB2 promoter region ChIP real-time PCR primers

82

CHAPTER 4

FOXP3 IS A NOVEL TRANSCRIPTION SUPPRESSOR FOR ONCOGENE SKP2 IN BREAST CANCER

4.1 Abstract

SKP2 is an F-box component in the E3 ubiquitin ligase SCF (SKP1-Cul1-

Fbox) complex (Nakayama and Nakayama, 2006). Although it is clear that SKP2

over-expression is involved in cell cycle dysregulation and the pathogenesis of

multiple malignancies (Nakayama et al., 2004; Pagano, 2004), the genetic

lesions for its upregulation is poorly understood. Here we show that mouse

mammary carcinoma in mice heterozygous for the Foxp3 mutation exhibits

increased Skp2 expression. Specific induction of FoxP3 represses SkP2

expression, while SiRNA silencing of the FOXP3 gene in human breast epithelial

cells results in elevated levels of SKP2 transcripts. FoxP3 directly binds and

represses the Skp2 promoter. Moreover, analysis of over two hundred primary breast cancer samples revealed an inverse correlation between FOXP3 and

SKP2 levels. Taken together, our data reveal that FOXP3 is a novel transcriptional repressor of SKP2.

83 4.2 Introduction

Cancer pathogenesis involves both inactivation of tumor suppressor genes and activation of oncogenes (Macdonald et al., 2004; Vogelstein and

Kinzler, 2002). One of the most fascinating aspects of cancer biology is the interaction between cancer suppressor genes and oncogenes. Most of these interactions are at post-translational levels. Proteins encoded by tumor suppressor genes can inactivate oncogenes. One of the most clearly studied cases is tumor suppressor Rb which inhibits the family member of oncogenes (Macdonald et al., 2004; Vogelstein and Kinzler, 2002). Conversely, oncogenes can overcome the tumor suppressor proteins. For example, SKP2 causes degradation of tumor suppressor FOXO (Huang et al., 2005), as well as

CDK inhibitors, such as p27 (Nakayama and Nakayama, 2006; Pagano, 2004).

It is less clear whether such antagonism exists at the transcriptional level.

However, a recent study suggests that FOXO may repress expression of cyclin

D (Schmidt et al., 2002), a well-known oncogene.

Overexpression of SKP2 was recently found in a significant portion of cancers (Nakayama and Nakayama, 2006). SKP2 is a component of the E3 ubiquitin ligase SCF with specificity for CDK inhibitor p27, and was found to be essential for G2/M exit in cell cycle (Nakayama et al., 2004; Pagano, 2004). The gene targeting knockout Skp2 in the mouse causes delayed animal growth and cellular polyploidy in several specific tissues, which are rescued by double knockout p27 in Skp2-/- mouse (Nakayama et al., 2000; Nakayama et al., 2004).

The increased expression of SKP2 has been shown in about half of breast

84 cancers, especially among those with early onset and poor prognosis (Radke et al., 2005; Signoretti et al., 2002; Sonoda et al., 2006; Traub et al., 2006). Genetic

lesions responsible for over expression of SKP2 in cancers remains to be defined,

though gene amplification of SKP2 has recently been shown in non-small cell

lung carcinomas (Zhu et al., 2004). The transcriptional regulation SKP2 has also

been analyzed recently (Imaki et al., 2003; Zhang and Wang, 2006). An intriguing

question is whether aberrant genetic alterations on some specific transcriptional

regulators of SKP2 could cause over-expression of SKP2 oncogene in cancers.

We have recently reported that a germline mutation of Foxp3, an X-linked

gene encoding a member of the forkhead transcription factor, resulted in a high

rate of spontaneous cancers and increased susceptibility to carcinogen in the

mouse. In addition, widespread deletion and somatic mutations of FOXP3 have

also been observed in breast cancers. Interestingly, our extensive studies

indicated that FOXP3 is a transcriptional repressor of the HER2/ErbB2

oncogene. Interestingly, FOXP3 also suppresses growth and induces cell death

of MCF-7, a breast cancer line without HER2/ErbB2 over-expression. These data

raise the possibility that FOXP3 may repress other potential oncogenes. Since

SKP2 plays an important role in the resistance to retinoid acid induced growth arrest in MCF7 cells (Dow et al., 2001), we explored the role of FoxP3 in the regulation of SKP2 in breast cancer. Here we reported that the genetic lesions of

Foxp3 cause increased expression of SKP2 in mammary tumors. Furthermore, we demonstrated that Skp2 was negatively regulated by Foxp3 at the

transcriptional level, and down-regulation of SKP2 by FoxP3 subsequently up-

85 regulates another tumor suppressor p27 that usually is targeted by SKP2 for degradation. Thus our data illustrates an interesting interplay between the oncogenes and tumor suppressor genes in cell cycle regulation, and indicates that loss-expression of FOXP3 is one genetic lesion resulting in increased expression of SKP2 in breast cancer cells.

4.3 Materials and Methods

Animals, Cell lines antibodies and quantitative real-time PCR

Antibodies for immunohistochemical staining and Western blotting: p27 (M-

197, 1:500, Santa Cruz), SKP2 (H-435, 1:100, Santa Cruz); Realtime PCR primer Skp2 Forward:TTAGTCGGGAGAACTTTCCAGGTG;

Skp2 Reverse: AGTCACGTCTGGGTGCAGATTT.

The rest of the materials are the same as Chapter 2 and 3,

Chromatin immunoprecipitation

This method was described in Chapter 3. The ChIP realtime PCR primers from upstream Skp2: ChIP-1 Forward: TGTGATGGGCACACATACAG; Reverse:

TGTTCTCTGGAAGCCTCAGC; ChIP-2 Forward: CGAATCTTGCTCTCTCCACA;

Reverse: CATGCAAAATTCAGGTGTGC; ChIP-3 Forward:

GGACAGGCTGTGGATTGAGT; Reverse: CCAAGAGGAGCGATGGTTTA;

ChIP-4 Forward: TGCTGGGACTTTTCTCCACT; Reverse:

AGACACCCATGCCTGATAGC.

86 FOXP3 silencing lentiviral vector

Same as Chapter 3.

Immunohistochemistry

Expression of FoxP3 and Skp2 in mouse and human breast cancer samples was determined using immunohistochemistry as described. The samples consisting of

>5% cells with nuclear expression of FoxP3 were scored as positive. SKP2 scoring followed the previous description (Radke et al., 2005; Signoretti et al.,

2002; Sonoda et al., 2006; Traub et al., 2006). Briefly, those with >5% either nuclear or cytoplasm positive staining accumulations were scored as Skp2+.

FOXP3 and SKP2 staining were scored double blind.

4.4 Results

4.4.1. Inactivation of FoxP3 locus results in increased SkP2 expression

We have used mice heterozygous for either Foxp3 alone or Foxp3 and

Otc genes to study the impact of the Foxp3 mutation on cancer susceptibility. As

demonstrated in Chapter 1, we showed that mice heterozygous for Foxp3 gene

alone developed spontaneous breast cancer at a high rate, while those

heterozygous for Otc gene showed no increase in cancer susceptibility.

Moreover, while most of the Foxp3 transcripts in the normal epithelial cells were

from wild-type alleles, all of the transcripts from the cancerous tissue were

transcribed from the mutant allele (Figure 2.5). Thus, in cancer cells, Foxp3 locus

is silenced. To determine whether Foxp3 represses Skp2 expression, first we

immunohistochemically stained the mammary tissue. As shown in Figure 4.1A,

87 Skp2 was found to be highly expressed in cancer cells compared with almost no expression in normal epithelial cells. To further quantify this increase of Skp2, we used real-time quantitative PCR. We captured cells from frozen sections by laser micro-dissection and extracted mRNA for RT-PCR analysis, which was described in Chapter 2. We compared expression of Skp2 in normal mammary epithelial cells from either WT or FoxP3sf/+Otcspf/+, as well as mammary cancer tissues from mutant mouse. As shown in Figure 4.1B, in comparison to WT epithelial cells, the heterozygous epithelial cells expressed a two-fold higher level of Skp2, which suggests a FoxP3 gene dose effect on the level of Skp2. Moreover, in cancerous tissue that has silenced the wild-type allele, expression of Skp2 was substantially enhanced. Thus, in vivo inactivation of Foxp3 locus resulted in increased Skp2 expression. As an alternative approach, we compared mouse mammary cancer tissues from WT and Foxp3 single mutant mice for expression of Skp2. As shown in Table 4.1, spontaneous cancers in WT mice had less Skp2 expression compared with those raised in Foxp3 mutant mouse.

To substantiate that inactivation of FOXP3 is a primary event leading to over-expression of SKP2, we transduced normal human mammary epithelial cells (HMEC) with lentiviral vector encoding siRNA specific for FOXP3. The un- transduced cells were eliminated by blasticidin. As shown in Chapter 3 Figure

3.4A, the FOXP3 siRNA transduction caused a more than 100-fold reduction in the FOXP3 transcript. Corresponding to this, a 4-fold increase of the SKP2 transcripts was observed (Figure 4.1 C).

88 4.4.2 The inverse correlation between FOXP3 and SKP2 in human breast

cancers.

As we showed in Chapter 1, most human breast cancers lose expression

of FOXP3, and widespread gene deletions and somatic mutations of FOXP3

were found in breast carcinomas. To explore if the loss-of-expression FOXP3 in

human breast cancer correlates with SKP2 up regulation, we did

immunohistochemical staining of FOXP3 and SKP2 in 206 human breast

carcinomas in tissue microassay slides. As shown in Figure 4.2, down-regulation

of FOXP3 was significantly associated with the over-expression of SKP2 in these

breast cancers. In conjunction with the observations of over expressed Skp2 in

mouse mammary tumors developed in Foxp3 mutant mice, the inverse

correlation between FOXP3 and SKP2 in human breast cancer suggests that

FOXP3 may be a repressor for SKP2.

4.4.3 FoxP3 as a transcriptional repressor of SkP2

Since we showed FOXP3 is a transcriptional repressor for another oncogene HER2/ErbB2 in Chapter 3, we carried out a similar approach to test whether this was the same case for SKP2. We transfected mouse mammary cancer line TSA with V5-targeted FoxP3 protein and selected two transfectant lines with stable Foxp3-V5 expression, referred to as Foxp3-V5 CL302 and

CL305. By using real-time PCR analysis, we found that Skp2 transcripts decreased by around 20-fold in both of the Foxp3-V5 transfectant lines compared with vector control, while there were no changes in p27 mRNA levels (Figure

4.3A). Since SkP2 targets p27 for degradation (Nakayama and Nakayama, 2006),

89 we also examined the protein levels of these two cell cycle regulators in Foxp3-

V5 transfectant CL302 and found that SKP2 was dramatically reduced by Foxp3, while p27 was significantly increased in the Foxp3-V5 transfectants.

To further confirm the down-regulation of Skp2 by Foxp3 occurring at the transcription level, we cloned the 2.0kb upstream of the murine Skp2 gene into the luciferase reporter vector pGL2 and tested the effects of Foxp3 on Skp2 promoter activity by luciferse assay. As shown in Figure 4.3B, FoxP3 substantially repressed the promoter activity of the Skp2 gene. Using a similar approach, Chromatin IP and quantitative real-time PCR, we found FoxP3 also binds the specific forkhead binding sites located around 1.0 kb upstream of the

Skp2 gene (Figure 4.4A). Site-directed mutagenesis of two binding sites at this region disrupts the repression of promoter activity by FoxP3, which suggests that

FoxP3 regulates Skp2 transcription through specific binding of this promoter region. Taken together, data in Figure 4.3 and 4.4 support the notion that FoxP3 is a transcriptional repressor of the Skp2 gene.

4.4.4 FoxP3 over-expression causes apoptosis and polyploidy of breast cancer cell lines

One of the clearly defined functions of Skp2 is to regulate the p27 levels at the G2/M phase in the cell cycle (Nakayama et al., 2004; Pagano, 2004), though a widely accepted function of p27 is the inhibition of G1-S transition in cell cycle

(Slingerland and Pagano, 2000). The compelling genetic evidence supports the critical function of Skp2 and p27 in G2-M progression from the elegant study of the double knockout p27 and Skp2 mouse model which sufficiently rescued most

90 of the phenotypes in the Skp2 single knockout mouse model, including retarding

growth, increasing apoptosis and marked polyploidy in several specific tissues

(Nakayama et al., 2004), which strongly suggests that p27 is the essential target

for SKP2 to fulfill its physiological function. As we found that down-regulation of

Skp2 by FoxP3 increases p27 protein level in Foxp3-V5 transfectant cells (Figure

4.3), we carried out the following experiments to ask whether Foxp3 could exert

physiological effects on cell cycle and death with the similar phenotype found in

Skp2-/- cells. First, we produced a pBI-EGFP-FOXP3 vector (Figure 4.5A) and

transfected it into the Tet-off MCF cell line. As shown in Figure 4.5 B upper

panels, in the absence of doxycycline, significant levels of FoxP3 were induced, and SKP2 transcript levels were subsequently reduced in a time dependent manner (Figure 4.5C). Correspondingly, after inducing FOXP3 expression, most

of the cells started to die by the fourth day of FOXP3 induction (Figure 4.5C

lower panel). Second, we compared the DNA contents of the FoxP3 high and

FoxP3 low subsets of the TSA-FoxP3-V5 cell line (CL302). As shown in Figure

4.6B, about 25% of the cells in the FoxP3 high subset had >4C of DNA contents.

In contrast, the FoxP3 low subset had relatively normal DNA contents. Thus, over-expression of FoxP3 resulted in polyploidy of cancer cells. Interestingly, the most FoxP3 low cells were hindered in the G1 phase and had less cells in the S

and G2 stages compared with vector control, which might indicate the effects of

p27 on blocking G1-S transition or the input from other unclear targets of FoxP3.

Taken together, we demonstrated that Foxp3 blocks the cell cycle and causes the polyploidy as well as eventually induces cell death.

91 4.5 Discussion

As a component of the SCF E3 complex, SKP2 has emerged as an

important oncogene (Nakayama and Nakayama, 2006). Over-expression of

SKP2 is associated with anpoor prognosis in breast cancer and is prevalent in

cancer that develops in young women (Signoretti et al., 2002). While over-

expression of Skp2 has been reported in multiple lineages of cancer and may

confer resistance to anti-steroid therapy (Nakayama and Nakayama, 2006;

Signoretti et al., 2002), the mechanism for its over-expression has not been well

understood. Here we demonstrate that FoxP3, which we have recently demonstrated to be the first X-linked cancer suppressor gene, is a novel transacting regulator for Skp2 expression. Our conclusion is based on three lines of evidence.

First, inactivation of the FoxP3 locus resulted in elevation of Skp2 expression. Mammary epithelial cells from mice heterozygous for the Foxp3 mutation have a two-fold higher level of Foxp3. More importantly, the spontaneous mammary cancer developed in Foxp3sf/+ mutant mice completely inactivated the locus and drastically increased the level of Skp2 transcripts.

Since the silencing of the FOXP3 gene by siRNA in human mammary epithelial cells resulted in a significant increase of the SKP2 transcript, it is likely that the elevation of Skp2 in mouse mammary cancers developed in the mice heterozygous for Foxp3 mutation is a direct consequence of disruption of Foxp3.

This is further supported by the fact that mammary cancers observed in wild-type

BALB/c mice showed much less Skp2 over-expression. The up-regulation of the

92 SKP2 gene was strongly associated with a lack of FOXP3 expression in 206 human breast cancers analyzed. Taken together, our data support the notion that the increase in SKP2 is likely a consequence of loss-of-expression of FOXP3 in breast cancer cells.

Second, transfection of FoxP3 into mammary cancer cell lines repressed the levels of Skp2 mRNA. Using Skp2 promoter driven luciferase reporter, we demonstrated that FoxP3 significantly repressed the promoter activity. Thus,

FoxP3 is a transcriptional repressor for the Skp2 gene. Moreover, chromatin immunoprecipitation assay demonstrated that FoxP3 binds to a specific region upstream of Skp2 gene, which was also confirmed by site-directed mutagenesis of FoxP3 binding sites in this region.

Thirdly, transfection of FoxP3 resulted in an increase of p27, the major target of SKP2. The increase of p27 was likely restricted to the G2/M phase as a high portion of FoxP3-expressing cells showed polyploidy, which was observed in the hepatocytes of the Skp2-deficient mice. Moreover, in the breast cancer line

MCF-7 Tet-off system, induction of FOXP3 caused apoptosis of the cells. The mechanism behind the increased apoptosis by FOXP3 remains to be defined, although a decrease of SKP2 by FOXP3 might be a possible reason for causing cell death since previous observations showed that down regulation of Skp2 caused apoptosis in a number of cell lines, including lung cancer and oral cancer cell lines (Nakayama and Nakayama, 2006).

Taken together, genetic, biochemical and functional analysis provided strong evidence that FoxP3 is a transcriptional repressor of the SKP2 oncogene.

93 Together with our previous data that FoxP3 is a transcriptional repressor of

Her2/ErbB2 oncogene, SKP2 is the second target for FoxP3. Identification of a

functionally important non-Her2 target explains why FoxP3 also repressed the

growth of MCF-7, a breast cancer cell line without HER2/ErbB2 over-expression.

It is worth noting that in non-small cell lung carcinomas, over-expression

of SKP2 was associated with the amplification of the SKP2 gene (Zhu et al.,

2004). It is therefore worth considering whether amplification of SKP2 conveys resistance of SKP2 expression to FoxP3-mediated inhibition. We consider it unlikely based on our analysis of HER2/ErbB2, another target of FOXP3 in breast cancer. We previously reported that regardless of HER2/ErbB2 gene copy numbers, defective expression of FOXP3 always led to a higher HER2 score in cancer tissues. Further studies are needed to determine whether

FOXP3 can attenuate SKP2 expression in cases with amplification of this oncogene.

94

Spontaneous Tumor Carcinogenesis induced Tumor**

Wild-type (n=5) Foxp3sf/+ (n=7) Wild-type (n=10) Foxp3sf/+ (n=12)

Skp2

(-) 4 2 7 4

(+) 1 5 3 8

*: The arabic numeral in the table indicates the number of mice carrying mammary tumors. **: DMBA was used to induce mammary tumorigenesis in mice, the detail protocol described in Chapter 2.4.1. Statistic analysis: Chi-square test P=0.0152

Table 4.1. Skp2 Expression in Mouse Mammary Tumors*

95 Figure 4.1 Inactivation of the FoxP3 locus results in increased SkP2

expression.

A. Immunohistochemistry staining of mammary cancers with an adjacent normal

mammary gland in one Foxp3sf/+ mouse shows Skp2 expression increased in cancer. B. Relative expression levels of Skp2 in normal mammary epithelium of

WT and Foxp3sf/+Otcspf/+ mice and the cancer tissues in the Foxp3sf/+Otcspf/+ mice,

as revealed by real-time RT-PCR of LCM samples. Data shown are means and

S.D. The expression of Skp2 was normalized against the internal control Hprt gene. Highly significant differences were observed between cancerous and normal tissue (P<0.001, ANOVA test when either internal standards were used).

C. Silencing of FOXP3 resulted in the up-regulation of SKP2 in primary human mammary epithelial cells (HMEC). FOXP3 was silenced in HMEC by using

siRNA that was previously described in Chapter 3 Figure 3.4A. SKP2 transcripts were quantified by real-time PCR. Data shown are means and SEM of relative levels of transcript (with that in the vector-transduced cells defined as 1.0) and represent those from three independent experiments (P<0.01, T test).

96

Figure 4.1 Inactivation of the FoxP3 locus results in increased SkP2 expression.

97

Figure 4.2 An inverse correlation between FOXP3 and SKP2 in human cancer samples.

By using either anti-FOXP3 antibody or anti-SKP2 antibody 206 human breast carcinomas in tissue microassay slides were immunohistochemically stained and were scored by two pathologists in a double blind fashion. Two representative cases and summary data from independent cases are presented. The P values of the chi-square tests are listed.

98 Figure 4.3 FoxP3 directly represses Skp2 and indirectly up-regulates p27

A. Transfection of Foxp3-V5 into TSA cells repressed expression of the Skp2 locus. The mRNA levels were measured by real-time PCR for the vector control cells and two stable Foxp3-V5 transfectant clones, CL302 and CL305.

Constitutive expression of Foxp3 in cells decreases Skp2 transcripts but not p27 mRNA level. Data shown are means and S.D. of triplicates. B. Foxp3 reduces

SKp2 and increases p27 proteins. Lysates of Foxp3-V5 or vector-transfected

TSA cell lines were analyzed by Western blot using anti-SKP2, anti-p27, anti-β-

Actin and anti-V5 antibodies. C. Foxp3 represses mouse Skp2 promotor activity.

The 2.0kb upstream of the mouse Skp2 gene was cloned into pGL2 luciferase reporter vector. Either Foxp3 cDNA or empty vector was transiently co- transfected with reporter vector at different ratios illustrated in the figure. The means of the vector group was artificially defined as 1.0. Mouse mammary cancer cell lines TSA was used for luciferase assay.

99

Figure 4.3 FoxP3 directly represses Skp2 and indirectly up-regulates p27

100 Figure 4.4 FoxP3 is a transcriptional repressor for Skp2 oncogene

A. The schematic depiction of the 5’ region of the Skp2 gene, including the

around 2.5kb upstream of exon 1 and partial intron 1. The forkhead binding

motifs are illustrated with small green bars, while the regions that were assayed

by real-time PCR are marked in red bars. The sonicated chromatins were

precipitated by either control IgG or anti-V5 mAb. The amounts of pull-down DNA

were determined relative to a standard curve generated from a titration of total

input DNA from the chromatin immuno-precipitation experiment (Im et al., 2004).

By subtraction of DNA precipitated by control IgG, the amount of DNA specifically precipitated by anti-V5 mAb is shown in the lower panel. B. Two forkhead binding

sites in the Skp2 promoter region were separately deleted and the deletion

sequences are; mut A: ACTAAACCAATATTCTAAT; Mut B:

TAAAAATAAACCATC). The mutations of either site prevent FOXP3-mediated

suppression. The promoter activity was measured and normalized as detailed in

Figure 4.3C, except that human breast cancer line T47D was used here.

101

Figure 4.4 FoxP3 is a transcriptional repressor for Skp2 oncogene

102 Figure 4.5 Inducing FOXP3 expression in Tet-off breast cancer cells causes cell death and reduces SKP2 expression

A. The schematic diagram shows the promoter structure in pBI-EGFP vector (BD

Biosciences Clontech, Cat#6154-1). The bi-directional promoter Pbi-1 is responsive to the tTA regulatory proteins in the Tet-Off System. Pbi-1 contains the tetracycline-responsive element (TRE) that is between two identical minimal

CMV promoters (PminCMV). EGFP is at one side while human FOXP3 cDNA was inserted at the PvuII/NheI sites in the opposite direction. B. Induction of FOXP3 causes cell death. The Tet-off MCF-7 line (BD Biosciences Clontech) was co- transfected with either pBI-EGFP-FOXP3 or pBI-EGFP control with pUB6 vector

(Invitrogen, Cat# V250-2) that conveys blasticidin resistance. The stable transfectants were selected by both blasticidin and G418 in doxycycline- containing medium. The stable transfectant cells were cultured in the absence of doxycycline for 5 days for induction of FOXP3. The cells were stained for either

FOXP3 or cell death markers (7AAD/Annexin-V) by flow cytometry on day three or day five respectively. Data shown are representative of those from two independent experiments. C. The figure shows FOXP3 kinetically repressed

SKP2 after FOXP3 being induced by withdraw doxycycline in medium.

103

Figure 4.5 Inducing FOXP3 expression in Tet-off breast cancer cells causes cell death and reduces SKP2 expression

104 Figure 4.6 FoxP3 blocks cell cycle and causes a polyploidy phenotype in mammary cancer cells

A. Histogram of flow cytometry shows FOXP3-V5 expression in TSA-vector cells and TSA-FOXP3V5 cells as analyzed by FITC-conjugated anti-V5 monoclonal antibody. B. Left panels show levels of FoxP3V5 fusion protein in vector control

(top) or FoxP3-transfected TSA cells (bottom), while the right panels show DNA contents of vector control (top), FoxP3hi (middle) and FoxP3lo cells. C. Summary of percentage of cells in different stages of the cell cycle. Compared with vector control cells, more cells were blocked in G1-0 phases in FoxP3lo cells, while more polyploidy cells were shown in FoxP3hi cells.

105

Figure 4.6 Foxp3 blocks cell cycle and causes a polyploidy phenotype in mammary cancer cells

106

CHAPTER 5

CONCLUDING REMARKS

To understand cancer is a long and tough journey for human beings. We

now know “cancer is, in essence, a genetic disease” (Vogelstein and Kinzler,

2002) through the tremendous efforts made by cancer researchers in the last

three decades. We have known that dominant proto-oncogenes are able to

transform normal cells to malignant ones, and that the germ-line mutations in certain tumor suppressor genes are the causes of hereditary cancer syndromes

in affected families (Alberts, 2002; Macdonald et al., 2004; Vogelstein and

Kinzler, 2002). However we still have a long way to go until we are able to claim

victory on the war on cancer, as most cancer patients are still suffering and the

cancer death rate has plateaued for the last five decades (ACS, 2006). Since the

genetic mechanisms of cancer genesis and growth is extremely complicated, it is

certain that there are far more cancer related genes yet to be discovered.

Using mutant mouse models to study cancer is one of the most promising

approaches to identify novel cancer related genes (Ghebranious and

Donehower, 1998; Hennighausen, 2000). Here we report that Foxp3sf/+ mutant mice develop multiple malignancies, especially mammary cancer, and we confirm the relevance of our findings to human cancer by showing that

107 widespread loss-of-expression, deletion and somatic mutations of FOXP3 in human breast cancer. BRCA1 and 2 are prototype breast cancer suppressor genes. However, only mammary-tissue-specific disruption of Brca1 and 2 genes in mice causes a high incidence of mammary cancer (Ludwig et al., 2001a;

Ludwig et al., 2001b; Xu et al., 2001; Xu et al., 1999). Interestingly, p53 -/+ mutant mice (Kuperwasser et al., 2000) in the BALB/c background have a similar mammary tumor phenotype as that of Foxp3sf/+ mutant mice in the same genetic background. This could be explained by the varying susceptibilities of different inbred mouse strains to mammary cancer, which have been repeatedly reported for p53 knockout mice (Donehower et al., 1992; Donehower et al., 1995). Thus it would be interesting to test if Foxp3 mutations in other inbred mouse strains could display the tumor spectrums similar to those that we found in BALB/c mice.

What role does the X chromosome play in breast cancer? How are X- linked genes involved in breast cancer? These questions attract more and more attention in the cancer genetic field (Spatz et al., 2004). Among few reports on these topics, David Livingston’s group (Ganesan et al., 2002; Richardson et al.,

2006) demonstrated the mechanism by which BRCA1 is involved in X inactivation in breast cancer cells and abnormality of X chromosome in some human breast cancers. Epidemiological studies (Kristiansen et al., 2005;

Kristiansen et al., 2002) of breast cancers in young female patients without

BRCA1 and 2 heritable mutations also revealed aberrant skewed X inactivation and indicate that the unknown X linked breast cancer suppressor genes are responsible for this tumor phenotype. Consistent with their findings, we showed

108 that the X-linked gene Foxp3 is silenced by skewed X inactivation in mammary tumors developed from Foxp3sf/+ mutant mice. In conjunction with our

observations of genetic changes at the FOXP3 locus on X chromosome in

human breast cancer, we identified FOXP3 as the first X-linked breast cancer

suppressor gene. Thus our study may help to accelerate research regarding the

relationship between the X chromosome and cancer.

Furthermore, we demonstrated that FOXP3 represses two critical

oncogenes in breast cancer, namely HER2/ErbB2 and SKP2. We first showed

that mouse ErbB2 and Skp2 expression increased in mammary tumors that

developed in Foxp3sf/+ mutant mice. We further demonstrated that FoxP3 binds

specific forkhead binding sites upstream of HER2/ErbB2 and SKP2 genes, and down-regulates their expression. To our knowledge, FoxP3 is the first forkhead transcription factor known to bind and regulate the promoter activities of

HER2/ErbB2 and SKP2 oncogenes. Thus, our data reveal an interesting interplay that exists between a tumor suppressor gene and oncogenes at the transcriptional level.

Taken together, we demonstrate that FoxP3 is the first X-linked breast cancer suppressor gene involved in the regulation of HER2/ErbB2 and SKP2 oncogenes in breast tumorigenesis. It is our sincere hope that these findings will open new vistas for both understanding and intervening breast cancer.

109

BIBLIOGRAPHY

Accili, D., and Arden, K. C. (2004). FoxOs at the crossroads of cellular metabolism, differentiation, and transformation. Cell 117, 421-426.

ACS (2006). Statistics for 2006 (American Cancer Society http://www.cancer.org/docroot/home/index.asp).

Alberts, B. (2002). Molecular biology of the cell (New York, Garland Science).

Aldaz, C. M., Liao, Q. Y., LaBate, M., and Johnston, D. A. (1996). Medroxyprogesterone acetate accelerates the development and increases the incidence of mouse mammary tumors induced by dimethylbenzanthracene. Carcinogenesis 17, 2069-2072.

Balmain, A. (2002). Cancer: new-age tumour suppressors. Nature 417, 235-237.

Banham, A. H., Beasley, N., Campo, E., Fernandez, P. L., Fidler, C., Gatter, K., Jones, M., Mason, D. Y., Prime, J. E., Trougouboff, P., et al. (2001). The FOXP1 winged helix transcription factor is a novel candidate tumor suppressor gene on chromosome 3p. Cancer Res 61, 8820-8829.

Bennett, C. L., Christie, J., Ramsdell, F., Brunkow, M. E., Ferguson, P. J., Whitesell, L., Kelly, T. E., Saulsbury, F. T., Chance, P. F., and Ochs, H. D. (2001). The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat Genet 27, 20-21.

Bofin, A. M., Ytterhus, B., Martin, C., O'Leary, J. J., and Hagmar, B. M. (2004). Detection and quantitation of HER-2 gene amplification and protein expression in breast carcinoma. Am J Clin Pathol 122, 110-119.

Boumil, R. M., and Lee, J. T. (2001). Forty years of decoding the silence in X- chromosome inactivation. Hum Mol Genet 10, 2225-2232.

110 Brodie, S. G., Xu, X., Qiao, W., Li, W. M., Cao, L., and Deng, C. X. (2001). Multiple genetic changes are associated with mammary tumorigenesis in Brca1 conditional knockout mice. Oncogene 20, 7514-7523.

Brown, C. J., Ballabio, A., Rupert, J. L., Lafreniere, R. G., Grompe, M., Tonlorenzi, R., and Willard, H. F. (1991). A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature 349, 38-44.

Brunkow, M. E., Jeffery, E. W., Hjerrild, K. A., Paeper, B., Clark, L. B., Yasayko, S. A., Wilkinson, J. E., Galas, D., Ziegler, S. F., and Ramsdell, F. (2001). Disruption of a new forkhead/winged-helix protein, scurfin, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat Genet 27, 68-73.

Burgering, B. M., and Kops, G. J. (2002). Cell cycle and death control: long live Forkheads. Trends Biochem Sci 27, 352-360.

Cardiff, R. D., Anver, M. R., Gusterson, B. A., Hennighausen, L., Jensen, R. A., Merino, M. J., Rehm, S., Russo, J., Tavassoli, F. A., Wakefield, L. M., et al. (2000). The mammary pathology of genetically engineered mice: the consensus report and recommendations from the Annapolis meeting. Oncogene 19, 968- 988.

Carlsson, P., and Mahlapuu, M. (2002). Forkhead transcription factors: key players in development and metabolism. Dev Biol 250, 1-23.

Carrel, L., and Willard, H. F. (2005). X-inactivation profile reveals extensive variability in X-linked gene expression in females. Nature 434, 400-404.

Cavenee, W. K., Dryja, T. P., Phillips, R. A., Benedict, W. F., Godbout, R., Gallie, B. L., Murphree, A. L., Strong, L. C., and White, R. L. (1983). Expression of recessive alleles by chromosomal mechanisms in retinoblastoma. Nature 305, 779-784.

Chang, X., Gao, J. X., Jiang, Q., Wen, J., Seifers, N., Su, L., Godfrey, V. L., Zuo, T., Zheng, P., and Liu, Y. (2005). The Scurfy mutation of FoxP3 in the thymus stroma leads to defective thymopoiesis. J Exp Med 202, 1141-1151.

Coffer, P. J., and Burgering, B. M. (2004). Forkhead-box transcription factors and their role in the . Nat Rev Immunol 4, 889-899.

111 Courtier, B., Heard, E., and Avner, P. (1995). Xce haplotypes show modified methylation in a region of the active X chromosome lying 3' to Xist. Proc Natl Acad Sci U S A 92, 3531-3535.

Cressman, V. L., Backlund, D. C., Hicks, E. M., Gowen, L. C., Godfrey, V., and Koller, B. H. (1999). Mammary tumor formation in p53- and BRCA1-deficient mice. Cell Growth Differ 10, 1-10.

Czech, M. P. (2003). Insulin's expanding control of forkheads. Proc Natl Acad Sci U S A 100, 11198-11200.

DeMars, R., LeVan, S. L., Trend, B. L., and Russell, L. B. (1976). Abnormal ornithine carbamoyltransferase in mice having the sparse-fur mutation. Proc Natl Acad Sci U S A 73, 1693-1697.

Donehower, L. A., Harvey, M., Slagle, B. L., McArthur, M. J., Montgomery, C. A., Jr., Butel, J. S., and Bradley, A. (1992). Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356, 215-221.

Donehower, L. A., Harvey, M., Vogel, H., McArthur, M. J., Montgomery, C. A., Jr., Park, S. H., Thompson, T., Ford, R. J., and Bradley, A. (1995). Effects of genetic background on tumorigenesis in p53-deficient mice. Mol Carcinog 14, 16-22.

Dow, R., Hendley, J., Pirkmaier, A., Musgrove, E. A., and Germain, D. (2001). Retinoic acid-mediated growth arrest requires ubiquitylation and degradation of the F-box protein Skp2. J Biol Chem 276, 45945-45951.

Dryja, T. P., Rapaport, J. M., Joyce, J. M., and Petersen, R. A. (1986). Molecular detection of deletions involving band q14 of chromosome 13 in retinoblastomas. Proc Natl Acad Sci U S A 83, 7391-7394.

Fogh, J., Wright, W. C., and Loveless, J. D. (1977). Absence of HeLa cell contamination in 169 cell lines derived from human tumors. J Natl Cancer Inst 58, 209-214.

Fontenot, J. D., Gavin, M. A., and Rudensky, A. Y. (2003). Foxp3 programs the development and function of CD4+CD25+ regulatory T cells. Nat Immunol 4, 330-336.

Fox, S. B., Brown, P., Han, C., Ashe, S., Leek, R. D., Harris, A. L., and Banham, A. H. (2004). Expression of the forkhead transcription factor FOXP1 is associated

112 with and improved survival in primary human breast carcinomas. Clin Cancer Res 10, 3521-3527.

Ganesan, S., Silver, D. P., Greenberg, R. A., Avni, D., Drapkin, R., Miron, A., Mok, S. C., Randrianarison, V., Brodie, S., Salstrom, J., et al. (2002). BRCA1 supports XIST RNA concentration on the inactive X chromosome. Cell 111, 393- 405.

Ghebranious, N., and Donehower, L. A. (1998). Mouse models in tumor suppression. Oncogene 17, 3385-3400.

Giovarelli, M., Musiani, P., Modesti, A., Dellabona, P., Casorati, G., Allione, A., Consalvo, M., Cavallo, F., di Pierro, F., De Giovanni, C., and et al. (1995). Local release of IL-10 by transfected mouse mammary adenocarcinoma cells does not suppress but enhances antitumor reaction and elicits a strong cytotoxic and antibody-dependent immune memory. J Immunol 155, 3112- 3123.

Godfrey, V. L., Rouse, B. T., and Wilkinson, J. E. (1994). Transplantation of T cell-mediated, lymphoreticular disease from the scurfy (sf) mouse. Am J Pathol 145, 281-286.

Groden, J., Thliveris, A., Samowitz, W., Carlson, M., Gelbert, L., Albertsen, H., Joslyn, G., Stevens, J., Spirio, L., Robertson, M., and et al. (1991). Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66, 589-600.

Guy, C. T., Webster, M. A., Schaller, M., Parsons, T. J., Cardiff, R. D., and Muller, W. J. (1992). Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease. Proc Natl Acad Sci U S A 89, 10578-10582.

Hanahan, D., and Weinberg, R. A. (2000). The hallmarks of cancer. Cell 100, 57- 70.

Harvey, M., McArthur, M. J., Montgomery, C. A., Jr., Bradley, A., and Donehower, L. A. (1993). Genetic background alters the spectrum of tumors that develop in p53-deficient mice. Faseb J 7, 938-943.

Heard, E. (2004). Recent advances in X-chromosome inactivation. Curr Opin Cell Biol 16, 247-255.

113 Hennighausen, L. (2000). Mouse models for breast cancer. Breast Cancer Res 2, 2-7.

Herman, J. G., Latif, F., Weng, Y., Lerman, M. I., Zbar, B., Liu, S., Samid, D., Duan, D. S., Gnarra, J. R., Linehan, W. M., and et al. (1994). Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma. Proc Natl Acad Sci U S A 91, 9700-9704.

Huang, H., Regan, K. M., Wang, F., Wang, D., Smith, D. I., van Deursen, J. M., and Tindall, D. J. (2005). Skp2 inhibits FOXO1 in tumor suppression through ubiquitin-mediated degradation. Proc Natl Acad Sci U S A 102, 1649-1654.

Hung, M. C., Matin, A., Zhang, Y., Xing, X., Sorgi, F., Huang, L., and Yu, D. (1995). HER-2/neu-targeting gene therapy--a review. Gene 159, 65-71.

Hurst, H. C. (2001). Update on HER-2 as a target for cancer therapy: the ERBB2 promoter and its exploitation for cancer treatment. Breast Cancer Res 3, 395- 398.

Huynh, K. D., and Lee, J. T. (2003). Inheritance of a pre-inactivated paternal X chromosome in early mouse embryos. Nature 426, 857-862.

Im, H., Grass, J. A., Johnson, K. D., Boyer, M. E., Wu, J., and Bresnick, E. H. (2004). Measurement of protein-DNA interactions in vivo by chromatin immunoprecipitation. Methods Mol Biol 284, 129-146.

Imaki, H., Nakayama, K., Delehouzee, S., Handa, H., Kitagawa, M., Kamura, T., and Nakayama, K. I. (2003). Cell cycle-dependent regulation of the Skp2 promoter by GA-binding protein. Cancer Res 63, 4607-4613.

Jimenez, R. E., Wallis, T., Tabasczka, P., and Visscher, D. W. (2000). Determination of Her-2/Neu status in breast carcinoma: comparative analysis of immunohistochemistry and fluorescent in situ hybridization. Mod Pathol 13, 37- 45.

Jones, P. A., and Laird, P. W. (1999). Cancer epigenetics comes of age. Nat Genet 21, 163-167.

Joslyn, G., Carlson, M., Thliveris, A., Albertsen, H., Gelbert, L., Samowitz, W., Groden, J., Stevens, J., Spirio, L., Robertson, M., and et al. (1991). Identification of deletion mutations and three new genes at the familial polyposis locus. Cell 66, 601-613.

114 Kaestner, K. H., Knochel, W., and Martinez, D. E. (2000). Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev 14, 142-146.

Kallioniemi, O. P., Kallioniemi, A., Kurisu, W., Thor, A., Chen, L. C., Smith, H. S., Waldman, F. M., Pinkel, D., and Gray, J. W. (1992). ERBB2 amplification in breast cancer analyzed by fluorescence in situ hybridization. Proc Natl Acad Sci U S A 89, 5321-5325.

Katoh, M. (2004). Human FOX gene family (Review). Int J Oncol 25, 1495-1500.

Knudson, A. G. (2001). Two genetic hits (more or less) to cancer. Nat Rev Cancer 1, 157-162.

Knudson, A. G., Jr. (1971). Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 68, 820-823.

Kraus, M. H., Popescu, N. C., Amsbaugh, S. C., and King, C. R. (1987). Overexpression of the EGF receptor-related proto-oncogene erbB-2 in human mammary tumor cell lines by different molecular mechanisms. Embo J 6, 605- 610.

Kristiansen, M., Knudsen, G. P., Maguire, P., Margolin, S., Pedersen, J., Lindblom, A., and Orstavik, K. H. (2005). High incidence of skewed X chromosome inactivation in young patients with familial non-BRCA1/BRCA2 breast cancer. J Med Genet 42, 877-880.

Kristiansen, M., Langerod, A., Knudsen, G. P., Weber, B. L., Borresen-Dale, A. L., and Orstavik, K. H. (2002). High frequency of skewed X inactivation in young breast cancer patients. J Med Genet 39, 30-33.

Kuperwasser, C., Hurlbut, G. D., Kittrell, F. S., Dickinson, E. S., Laucirica, R., Medina, D., Naber, S. P., and Jerry, D. J. (2000). Development of spontaneous mammary tumors in BALB/c p53 heterozygous mice. A model for Li-Fraumeni syndrome. Am J Pathol 157, 2151-2159.

Lai, E., Prezioso, V. R., Tao, W. F., Chen, W. S., and Darnell, J. E., Jr. (1991). Hepatocyte nuclear factor 3 alpha belongs to a gene family in mammals that is homologous to the Drosophila homeotic gene fork head. Genes Dev 5, 416-427.

Lee, J. T. (2005). Regulation of X-chromosome counting by Tsix and Xite sequences. Science 309, 768-771.

115 Lee, J. T., Davidow, L. S., and Warshawsky, D. (1999). Tsix, a gene antisense to Xist at the X-inactivation centre. Nat Genet 21, 400-404.

Lee, J. T., Strauss, W. M., Dausman, J. A., and Jaenisch, R. (1996). A 450 kb transgene displays properties of the mammalian X-inactivation center. Cell 86, 83-94.

Li, J., and Vogt, P. K. (1993). The retroviral oncogene qin belongs to the transcription factor family that includes the homeotic gene fork head. Proc Natl Acad Sci U S A 90, 4490-4494.

Li, S., Weidenfeld, J., and Morrisey, E. E. (2004). Transcriptional and DNA binding activity of the Foxp1/2/4 family is modulated by heterotypic and homotypic protein interactions. Mol Cell Biol 24, 809-822.

Ludwig, T., Fisher, P., Ganesan, S., and Efstratiadis, A. (2001a). Tumorigenesis in mice carrying a truncating Brca1 mutation. Genes Dev 15, 1188-1193.

Ludwig, T., Fisher, P., Murty, V., and Efstratiadis, A. (2001b). Development of mammary adenocarcinomas by tissue-specific knockout of Brca2 in mice. Oncogene 20, 3937-3948.

Lydon, J. P., Ge, G., Kittrell, F. S., Medina, D., and O'Malley, B. W. (1999). Murine mammary gland carcinogenesis is critically dependent on progesterone receptor function. Cancer Res 59, 4276-4284.

Lyon, M. F. (1961). GENE ACTION IN X-CHROMOSOME OF MOUSE (MUS MUSCULUS L). NATURE 190, 372-373.

Macdonald, F., Ford, C. H. J., and Casson, A. G. (2004). Molecular biology of cancer, 2nd edn (London ; New York, N.Y., BIOS Scientific Publishers).

Mak, W., Nesterova, T. B., de Napoles, M., Appanah, R., Yamanaka, S., Otte, A. P., and Brockdorff, N. (2004). Reactivation of the paternal X chromosome in early mouse embryos. Science 303, 666-669.

Mayo, M. W., Wang, C. Y., Drouin, S. S., Madrid, L. V., Marshall, A. F., Reed, J. C., Weissman, B. E., and Baldwin, A. S. (1999). WT1 modulates apoptosis by transcriptionally upregulating the bcl-2 proto-oncogene. Embo J 18, 3990-4003.

Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P. A., Harshman, K., Tavtigian, S., Liu, Q., Cochran, C., Bennett, L. M., Ding, W., and et al. (1994). A 116 strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 266, 66-71.

Muller, W. J., Sinn, E., Pattengale, P. K., Wallace, R., and Leder, P. (1988). Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 54, 105-115.

Nakayama, K., Nagahama, H., Minamishima, Y. A., Matsumoto, M., Nakamichi, I., Kitagawa, K., Shirane, M., Tsunematsu, R., Tsukiyama, T., Ishida, N., et al. (2000). Targeted disruption of Skp2 results in accumulation of cyclin E and p27(Kip1), polyploidy and centrosome overduplication. Embo J 19, 2069-2081.

Nakayama, K., Nagahama, H., Minamishima, Y. A., Miyake, S., Ishida, N., Hatakeyama, S., Kitagawa, M., Iemura, S., Natsume, T., and Nakayama, K. I. (2004). Skp2-mediated degradation of p27 regulates progression into mitosis. Dev Cell 6, 661-672.

Nakayama, K. I., and Nakayama, K. (2006). Ubiquitin ligases: cell-cycle control and cancer. Nat Rev Cancer 6, 369-381.

Okamoto, I., Otte, A. P., Allis, C. D., Reinberg, D., and Heard, E. (2004). Epigenetic dynamics of imprinted X inactivation during early mouse development. Science 303, 644-649.

Pagano, M. (2004). Control of DNA synthesis and mitosis by the Skp2-p27- Cdk1/2 axis. Mol Cell 14, 414-416.

Parolini, O., Ressmann, G., Haas, O. A., Pawlowsky, J., Gadner, H., Knapp, W., and Holter, W. (1998). X-linked Wiskott-Aldrich syndrome in a girl. N Engl J Med 338, 291-295.

Pietenpol, J. A., Bohlander, S. K., Sato, Y., Papadopoulos, N., Liu, B., Friedman, C., Trask, B. J., Roberts, J. M., Kinzler, K. W., Rowley, J. D., and et al. (1995). Assignment of the human p27Kip1 gene to 12p13 and its analysis in . Cancer Res 55, 1206-1210.

Radke, S., Pirkmaier, A., and Germain, D. (2005). Differential expression of the F-box proteins Skp2 and Skp2B in breast cancer. Oncogene 24, 3448-3458.

Ramsdell, F., and Ziegler, S. F. (2003). Transcription factors in . Curr Opin Immunol 15, 718-724.

117 Richardson, A. L., Wang, Z. C., De Nicolo, A., Lu, X., Brown, M., Miron, A., Liao, X., Iglehart, J. D., Livingston, D. M., and Ganesan, S. (2006). X chromosomal abnormalities in basal-like human breast cancer. Cancer Cell 9, 121-132.

Schmidt, M., Fernandez de Mattos, S., van der Horst, A., Klompmaker, R., Kops, G. J., Lam, E. W., Burgering, B. M., and Medema, R. H. (2002). Cell cycle inhibition by FoxO forkhead transcription factors involves downregulation of cyclin D. Mol Cell Biol 22, 7842-7852.

Schubert, L. A., Jeffery, E., Zhang, Y., Ramsdell, F., and Ziegler, S. F. (2001). Scurfin (FOXP3) acts as a repressor of transcription and regulates T cell activation. J Biol Chem 276, 37672-37679.

Scott, G. K., Chang, C. H., Erny, K. M., Xu, F., Fredericks, W. J., Rauscher, F. J., 3rd, Thor, A. D., and Benz, C. C. (2000). Ets regulation of the erbB2 promoter. Oncogene 19, 6490-6502.

Shih, C., Shilo, B. Z., Goldfarb, M. P., Dannenberg, A., and Weinberg, R. A. (1979). Passage of phenotypes of chemically transformed cells via transfection of DNA and chromatin. Proc Natl Acad Sci U S A 76, 5714-5718.

Shu, W., Yang, H., Zhang, L., Lu, M. M., and Morrisey, E. E. (2001). Characterization of a new subfamily of winged-helix/forkhead (Fox) genes that are expressed in the lung and act as transcriptional repressors. J Biol Chem 276, 27488-27497.

Signoretti, S., Di Marcotullio, L., Richardson, A., Ramaswamy, S., Isaac, B., Rue, M., Monti, F., Loda, M., and Pagano, M. (2002). Oncogenic role of the ubiquitin ligase subunit Skp2 in human breast cancer. J Clin Invest 110, 633-641.

Sirchia, S. M., Ramoscelli, L., Grati, F. R., Barbera, F., Coradini, D., Rossella, F., Porta, G., Lesma, E., Ruggeri, A., Radice, P., et al. (2005). Loss of the inactive X chromosome and replication of the active X in BRCA1-defective and wild-type breast cancer cells. Cancer Res 65, 2139-2146.

Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., and McGuire, W. L. (1987). Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235, 177-182.

Slamon, D. J., Leyland-Jones, B., Shak, S., Fuchs, H., Paton, V., Bajamonde, A., Fleming, T., Eiermann, W., Wolter, J., Pegram, M., et al. (2001). Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344, 783-792. 118 Slingerland, J., and Pagano, M. (2000). Regulation of the cdk inhibitor p27 and its deregulation in cancer. J Cell Physiol 183, 10-17.

So, C. W., and Cleary, M. L. (2004). Dimerization: a versatile switch for oncogenesis. Blood 104, 919-922.

Sonoda, H., Inoue, H., Ogawa, K., Utsunomiya, T., Masuda, T. A., and Mori, M. (2006). Significance of skp2 expression in primary breast cancer. Clin Cancer Res 12, 1215-1220.

Soule, H. D., Vazguez, J., Long, A., Albert, S., and Brennan, M. (1973). A human cell line from a pleural effusion derived from a breast carcinoma. J Natl Cancer Inst 51, 1409-1416.

Spatz, A., Borg, C., and Feunteun, J. (2004). X-chromosome genetics and human cancer. Nat Rev Cancer 4, 617-629.

Stehelin, D., Varmus, H. E., Bishop, J. M., and Vogt, P. K. (1976). DNA related to the transforming gene(s) of avian sarcoma viruses is present in normal avian DNA. Nature 260, 170-173.

Tommasini, A., Ferrari, S., Moratto, D., Badolato, R., Boniotto, M., Pirulli, D., Notarangelo, L. D., and Andolina, M. (2002). X-chromosome inactivation analysis in a female carrier of FOXP3 mutation. Clin Exp Immunol 130, 127-130.

Traub, F., Mengel, M., Luck, H. J., Kreipe, H. H., and von Wasielewski, R. (2006). Prognostic impact of Skp2 and p27 in human breast cancer. Breast Cancer Res Treat.

van de Vijver, M. J., Peterse, J. L., Mooi, W. J., Wisman, P., Lomans, J., Dalesio, O., and Nusse, R. (1988). Neu-protein overexpression in breast cancer. Association with comedo-type ductal carcinoma in situ and limited prognostic value in stage II breast cancer. N Engl J Med 319, 1239-1245.

Venkatachalam, S., and Donehower, L. A. (1998). Murine tumor suppressor models. Mutat Res 400, 391-407.

Veres, G., Gibbs, R. A., Scherer, S. E., and Caskey, C. T. (1987). The molecular basis of the sparse fur mouse mutation. Science 237, 415-417.

Vogelstein, B., and Kinzler, K. W. (2002). The genetic basis of human cancer, 2nd edn (New York ; London, McGraw-Hill). 119 Wang, B., Lin, D., Li, C., and Tucker, P. (2003). Multiple domains define the expression and regulatory properties of Foxp1 forkhead transcriptional repressors. J Biol Chem 278, 24259-24268.

Weigel, D., Jurgens, G., Kuttner, F., Seifert, E., and Jackle, H. (1989). The homeotic gene fork head encodes a nuclear protein and is expressed in the terminal regions of the Drosophila embryo. Cell 57, 645-658.

Wildin, R. S., Ramsdell, F., Peake, J., Faravelli, F., Casanova, J. L., Buist, N., Levy-Lahad, E., Mazzella, M., Goulet, O., Perroni, L., et al. (2001). X-linked neonatal mellitus, enteropathy and endocrinopathy syndrome is the human equivalent of mouse scurfy. Nat Genet 27, 18-20.

Wooster, R., Bignell, G., Lancaster, J., Swift, S., Seal, S., Mangion, J., Collins, N., Gregory, S., Gumbs, C., and Micklem, G. (1995). Identification of the breast cancer susceptibility gene BRCA2. Nature 378, 789-792.

Wooster, R., and Weber, B. L. (2003). Breast and ovarian cancer. N Engl J Med 348, 2339-2347.

Xia, S. J., Pressey, J. G., and Barr, F. G. (2002). Molecular pathogenesis of rhabdomyosarcoma. Cancer Biol Ther 1, 97-104.

Xing, X., Wang, S. C., Xia, W., Zou, Y., Shao, R., Kwong, K. Y., Yu, Z., Zhang, S., Miller, S., Huang, L., and Hung, M. C. (2000). The ets protein PEA3 suppresses HER-2/neu overexpression and inhibits tumorigenesis. Nat Med 6, 189-195.

Xu, N., Tsai, C. L., and Lee, J. T. (2006). Transient homologous chromosome pairing marks the onset of X inactivation. Science 311, 1149-1152.

Xu, X., Qiao, W., Linke, S. P., Cao, L., Li, W. M., Furth, P. A., Harris, C. C., and Deng, C. X. (2001). Genetic interactions between tumor suppressors Brca1 and p53 in apoptosis, cell cycle and tumorigenesis. Nat Genet 28, 266-271.

Xu, X., Wagner, K. U., Larson, D., Weaver, Z., Li, C., Ried, T., Hennighausen, L., Wynshaw-Boris, A., and Deng, C. X. (1999). Conditional mutation of Brca1 in mammary epithelial cells results in blunted ductal morphogenesis and tumour formation. Nat Genet 22, 37-43.

Yarden, Y., and Sliwkowski, M. X. (2001). Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2, 127-137.

120 Yaziji, H., Goldstein, L. C., Barry, T. S., Werling, R., Hwang, H., Ellis, G. K., Gralow, J. R., Livingston, R. B., and Gown, A. M. (2004). HER-2 testing in breast cancer using parallel tissue-based methods. Jama 291, 1972-1977.

Ye, X., Robinson, M. B., Batshaw, M. L., Furth, E. E., Smith, I., and Wilson, J. M. (1996). Prolonged metabolic correction in adult ornithine transcarbamylase- deficient mice with adenoviral vectors. J Biol Chem 271, 3639-3646.

Zhang, L., and Wang, C. (2006). F-box protein Skp2: a novel transcriptional target of E2F. Oncogene 25, 2615-2627.

Zhu, C. Q., Blackhall, F. H., Pintilie, M., Iyengar, P., Liu, N., Ho, J., Chomiak, T., Lau, D., Winton, T., Shepherd, F. A., and Tsao, M. S. (2004). Skp2 gene copy number aberrations are common in non-small cell lung carcinoma, and its overexpression in tumors with ras mutation is a poor prognostic marker. Clin Cancer Res 10, 1984-1991.

Ziegler, S. F. (2006). FOXP3: Of Mice and Men. Annu Rev Immunol 24, 209-226.

121