Characterizing the Function of the Fps/Fes Tyrosine Kinase

in the Mammary Gland

by

PETER FRANCIS TRUESDELL

A thesis submitted to the Department of Pathology and Molecular Medicine

In conformity with the requirements for

the degree of Doctor of Philosophy

Queen’s University

Kingston, Ontario, Canada

June, 2008

Copyright © Peter Francis Truesdell, 2008

ii

Abstract

The fps proto-oncogene encodes a 92 kDa cytoplasmic tyrosine kinase. Previous

studies have shown that Fps expression in the mammary gland changes with

development, and Fps has a suppressor function in mammary tumorigenesis. The aim of

my thesis was to elucidate the role of the Fps tyrosine kinase in regulating mammary

gland development and function. We have shown that the expression of the Fps kinase in

the mammary gland increased during pregnancy and reached its maximum during

lactation. The level of Fps tyrosine phosphorylation paralleled the expression pattern.

Pups reared by fps-null females gained weight more slowly than those reared by wild-

type females. Epithelial cells were the primary source of Fps expression. Milk protein

and fat content were not affected by the absence of Fps. Similarly, no differences in

mammary gland structure were observed with whole mount or histological analysis.

Fps was shown to be in a multi-protein complex with E-cadherin, β-catenin and

p120-catenin. A strong co-localization signal was observed for Fps and E-cadherin.

Immunofluorescence analysis indicated that the localization of E-cadherin and β-catenin

was disorganized and less concentrated at sites of cell-cell contacts in the fps-null glands.

The interactions between the different adherens junction components were altered in the

fps-null tissue. Specifically, less E-cadherin and β-catenin was associated with p120-

catenin in the fps-null glands. Suprisingly, no phosphotyrosine differences were detected

for the adherens junction components.

Conditions were established to grow primary murine epithelial cell cultures that

could be used to investigate the function of Fps. Fps expression was up-regulated in these

iii

cells in response to lactogenic hormones. A lentiviral system encoding a murine p53

shRNA sequence was used to increase the growth potential of the primary cells.

Continual growth of the infected and uninfected primary epithelial cell mixture resulted

in the establishment of an immortalized cell line. Immunofluorescent and immunoblot

analyses revealed that the cells have undergone an epithelial-to-mesenchymal transition.

With the transduction of a myc-epitope tagged Fps into the cells, we have generated cell

lines with the appropriate genetic backgrounds to study the function of the Fps kinase in

the mammary gland, specifically as it relates to tumorigenesis.

iv

Co-Authorship

The co-authors of this thesis are Waheed Sangrar, Yan Gao, Ralph Zirngibl, Sarah

Francis and Peter Greer. The contributions that these authors made to the manuscripts are

listed below.

Chapter 2: The authors are Peter Truesdell, Ralph Zirngibl, Sarah Francis, Waheed

Sangrar and Peter Greer. I designed and performed the majority of the experiments, with

the exception of Figure 3 which was produced by Ralph Zirngibl. Sarah Francis and

Waheed Sangrar contributed to Figure 8. I wrote the manuscript. Editing was done by

Peter Greer.

Chapter 3: The authors are Peter Truesdell, Yan Gao, Waheed Sangrar and Peter Greer.

I designed and performed the majority of the experiments. Yan Gao and Waheed Sangrar

contributed to the isolation of primary epithelial cells and confocal data (Figure 1 and 5).

I wrote the manuscript. Editing was done by Peter Greer.

v

Acknowledgements

I would like to thank my supervisor, Dr. Peter Greer, for allowing me the

opportunity to study in his lab. His guidance, patience and expertise have been

instrumental to the success of this project. His dedication and support for the lab is never

in question.

I would also like to thank current and past members of the lab for helpful

discussions and making the lab an enjoyable place to work. A thank you also goes to the

Ontario Graduate Scholarship, the Queen’s Graduate Award and the United States Army

Breast Cancer Research Program for generously funding this research project.

vi

Table of Contents

Abstract ............................................................................................................................... ii

Co-Authorship.................................................................................................................... iv

Acknowledgements ............................................................................................................. v

Table of Contents ............................................................................................................... vi

List of Figures ..................................................................................................................... x

List of Abbreviations ....................................................................................................... xiii

Chapter 1: General Introduction……...……………………………………………….......1

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

1.2 The Fps and Fer tyrosine kinases.............................................................................3

1.2.1 The discovery of fps/fes as a component of retroviral oncogenes...........................3

1.2.2 The fps and fer protein-tyrosine kinase-encoding proto-oncogenes........................6

1.2.3 Expression patterns and subcellular localization of Fps and Fer tyrosine

kinases......................................................................................................................9

1.2.4 Structural components and kinase activity of the Fps and Fer tyrosine kinases....12

1.2.5 Transgenic mouse models of the Fps tyrosine kinase............................................19

1.2.6 The Fps tyrosine kinase in the mammary gland........................ ………................23

1.3 The mammary gland..............................................................................................24

1.3.1 Mammary gland development...............................................................................26

1.3.2 Early mammary gland development......................................................................26

1.3.3 The mammary gland during pregnancy and lactation...........................................28

1.3.4 Involution..............................................................................................................33

vii

1.4 Milk........................................................................................................................36

1.4.1 The secretion of milk components.........................................................................36

1.4.2 The exocytotic and lipid secretion pathways.........................................................38

1.5 Cell adhesion..........................................................................................................41

1.5.1 Cadherin-mediated cell adhesion...........................................................................41

1.5.2 Targeted disruption of adherens junction components..........................................47

1.5.3 Cadherins, catenins and cancer..............................................................................49

1.5.4 Regulation of the adherens junction by tyrosine phosphorylation........................53

1.6 Hypothesis and research objectives.......................................................................57

Chapter 2: The Fps tyrosine kinase is up-regulated during lactation in the mouse

mammary gland and is a component of the E-cadherin adherens junction.

2.1 Abstract………………………………………………………………………….59

2.2 Introduction……………………………………………………………………...60

2.3 Materials and Methods ………………………………………………………….66

2.3.1 Mice……………………………………………………………………………...66

2.3.2 Pup weight analysis………………………………………………………….…..67

2.3.3 Antibodies…………………………………………………………………….…67

2.3.4 Immunoblotting, immunoprecipitation and kinase assay analysis………………68

2.3.5 Whole mount and histological analysis………………………………………….69

2.3.6 Immunofluorescence……………………………………………………….……70

2.3.7 Milk protein and fat analysis…………………………………………………….71

2.4 Results…………………………………………………………………………...71

viii

2.4.1 Expression of the Fps tyrosine kinase increases during pregnancy and reaches

maximal levels during lactation………………………………………………….71

2.4.2 The Fps tyrosine kinase is highly phosphorylated during lactation……………..72

2.4.3 Offspring raised by fps-null females gained weight more slowly than those raised

by wild-type females……………………………………………………….……76

2.4.4 The Fps tyrosine kinase is expressed in mammary epithelial cells……………...78

2.4.5 Milk from fps-null mice has normal protein and fat content…………………....78

2.4.6 Analysis of the expression and phosphotyrosine content of various signaling

molecules in the mammary gland………………………………………………..81

2.4.7 The morphology of the fps-null mammary gland is normal……………………..81

2.4.8 The Fps and Fer tyrosine kinases are components of the E-cadherin-based

adherens junction in the mammary gland during lactation………………………85

2.4.9 Co-localization of the Fps tyrosine kinase with E-cadherin in epithelial cells

during lactation and altered E-cadherin and β-catenin staining in fps-null

mammary glands…………………………………………………………………88

2.4.10 Protein-protein interactions between the components of the adherens junction...90

2.4.11 Endogenous phosphotyrosine levels of AJ components………………….……...90

2.4.12 Fps and Fer kinase activities are associated with p120-catenin……………….....93

2.5 Discussion………………………………………………………………………..95

Chapter 3: Immortalization of murine mammary epithelial cells by targeting p53

expression with a lentivirus-mediated RNA interference system.

3.1 Abstract………………………………………………………………………...106

3.2 Introduction………………………………………………………………….....107

ix

3.3 Materials and Methods…………………………………………………………110

3.3.1 Materials………………………………………………………………………..110

3.3.2 Antibodies……………………………………………………………………....111

3.3.3 Lactogenic hormone induction ………………………………………………...111

3.3.4 Generation of p53 RNAi lentivirus and MSCVpac/Fpsmyc virus……………..111

3.3.5 Transgenic mice and isolation of epithelial cells……………………………….112

3.3.6 Immunofluorescence……………………………………………………………114

3.3.7 Western analysis………………………………………………………………..115

3.4 Results…………………………………………………………………………..116

3.5 Discussion………………………………………………………………………127

Chapter 4: General Discussion…………………………………………....…………….135

References………………………………………………………………..........………..153

x

List of Figures

Figure 1.1 Cytoplasmic protein tyrosine kinases..........................................................2

Figure 1.2 Comparison of cellular Fps and Fer kinases and viral Fps oncoproteins…4

Figure 1.3 Organization of the fps and fer loci……………………………………….7

Figure 1.4 Cellular components of the mouse mammary gland…………………….25

Figure 1.5 Mammary epithelial cell exocytotic and lipid secretion pathways that

contribute to the production of milk during lactation……………………37

Figure 1.6 Structural components of the E-cadherin-based adherens junction……...43

Figure 2.1 Expression of the Fps tyrosine kinase during mammary development.....73

Figure 2.2 The Fps tyrosine kinase is highly activated in the mammary gland during

lactation…………………………………………………………………..74

Figure 2.3 Pups reared by fps-null mothers gain weight more slowly than wild-type

mothers…………………………………………………………………...77

Figure 2.4 Cell-specific expression of the Fps tyrosine kinase in the mammary gland

during lactation………………………………………………………......79

Figure 2.5 Analysis of milk proteins and triglycerides……………………………...80

Figure 2.6 Phosphotyrosine analyses of potential substrates of the Fps tyrosine

kinase…………………………………………………………………….82

Figure 2.7 Whole mount and histological analysis of wild-type and fps-null

mammary glands……………………………………………………...83-84

Figure 2.8 The Fps tyrosine kinase is a component of the E-cadherin-based adherens

junction in the mammary gland during lactation……………………..86-87

xi

Figure 2.9 Co-localization of the Fps tyrosine kinase with E-cadherin in the lactating

mammary gland and the subcellular localization of E-cadherin and β-

catenin in glands from fps-null mice……………………………………..89

Figure 2.10 Differences in the interactions between the adherens junction components

from wild-type and fps-null mammary glands ………………………......91

Figure 2.11 Endogenous phosphotyrosine content of the adherens junction

components…………………………………………………………………..….92

Figure 2.12 In vitro kinase assays of the adherens junction components……………...........94

Figure 3.1 Growth of primary murine epithelial cells in a collagen gel matrix and on

collagen-coated surfaces……………………………………..................117

Figure 3.2 Induction of Fps expression in wild-type epithelial cells in response to

treatment with lactogenic hormones……………………………………118

Figure 3.3 Schematic representation of the lentiviral p53 shRNA construct used to

target p53 expression…………………………………………………...120

Figure 3.4 Enhanced survival of primary mammary epithelial cells infected with the

p53 shRNA lentivirus…………………………………………………..121

Figure 3.5 Retention of epithelial cell properties during the establishment of the fps-

null mammary cell line………………………………………………....122

Figure 3.6 Growth of epithelial cells at different steps of immortalization………..124

Figure 3.7 Expression of epithelial and non-epithelial markers in the fps-null

immortalized cells………………………………………………………125

Figure 3.8 Confocal immunofluorescence microscopy analysis of the fps-null

immortalized cells………………………………………………………128

xii

Figure 4.1 Proposed model for the role of Fps in E-cadherin-based AJ in the mouse

mammary gland during lactation……………………………………….150

xiii

List of Abbreviations

α alpha

ab antibody

ADPH adipophilin

AJ adherens junction

ATP adenosine triphosphate

β beta

BAR Bin/Amphiphysin/Rvs

BCR B cell receptor

BSA bovine serum albumin

Btn butyrophilin

CC coiled-coil

C/EBPδ CCAAT/enhancer binding protein delta

CIP4 Cdc42-interacting protein 4

CLD cytoplasmic lipid droplet

C-terminal carboxy terminal

DNA deoxyribonucleic acid

DNase deoxyribonuclease

E-cadherin Epithelial-cadherin

EC extracellular

EFC extended FCH domain

EGF epidermal growth factor

ELF E74-like factor

EMT epithelial-to-mesenchymal transition

ER estrogen receptor

Erk extracellular signal-regulated kinase

F-BAR FCH-containing BAR-like

FBP17 formin-binding protein 17

FBS fetal bovine serum

xiv

Fer Fes related

Fes Feline sarcoma

Fps Fujinami poultry sarcoma

FSV Fujinami sarcoma virus

GA Golgi apparatus

GAP GTPase activating protein

GFP green fluorescent protein

GH growth hormone

GLUT1 glucose transporter

GM-CSF granulocyte-macrophage-colony stimulating factor

GSK-3 glucose synthase kinase 3

GTP guanosine triphosphate

H&E hematoxylin and eosin

HGF hepatocyte growth factor

hr hour

IF intermediate filament

IgG immunoglobin G

IGF-2 insulin-like growth factor 2

IL interleukin

IP immunoprecipitation

IU international unit

JAK Janus kinase

Kb kilobase

kDa kilodalton

KRB kinase reaction buffer

LCR locus control region

LIF leukemia inhibitory factor

MAPK mitogen activated protein kinase

MF myristoylated Fps

MLD milk lipid drop

xv

MLG milk lipid globule

MMTV mouse mammary tumor virus

N-cadherin neural-cadherin

NF1-C2 nuclear factor 1-C2

NFkappaB nuclear factor kappaB

NSF N-ethylmaleimide-sensitive factor

N-terminal amino-terminal

O/N overnight

PBS phosphate buffered saline

PDGF platelet derived growth factor

PGK phosphoglycerate kinase

PI(3)K phosphoinositide-3 kinase

PIP phosphatidylinositol phosphate

PI(4,5)P2 phosphatidylinositol 4,5-bisphosphate

PR progesterone receptor

PTHrP parathyroid hormone-related peptide

PTK protein tyrosine kinase

PTP1B protein tyrosine phosphatase

pY phosphotyrosine

PyVmT polyoma virus middle T antigen

RANKL receptor activator of nuclear factor-kappa B ligand

RER rough endoplasmic reticulum

RNAi ribonucleic acid interference

RNase ribonuclease

RT room temperature

SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis

SER smooth endoplasmic reticulum

SH2 Src homology 2

SMA smooth muscle actin

SNAP soluble NSF attachment protein

xvi

SNARE soluble NSF attachment protein receptors

SREBP1 sterol regulatory element binding protein 1

Stat signal transducer and activator of transcription

SV secretory vesicle

TBST Tris-buffered saline Tween-20

TEB terminal end bud

TGFβ3 transforming growth factor beta

TGN trans Golgi network

TNF tumor necrosis factor

VEGF vascular endothelial growth factor

WAP whey acidic protein

WASP Wiskott-Aldrich syndrome protein

XOR xanthine oxidoreductase

1

Chapter 1

General introduction

1.1 Introduction

Protein kinases constitute one of the largest mammalian protein superfamilies and

are important regulators of cellular processes including cell cycle progression, cell

proliferation and apoptosis, gene expression, metabolism, motility, morphology and

differentiation. Protein kinases are defined by a conserved catalytic domain. Through the

phosphorylation of substrate proteins, kinases have the ability to alter the function of a

large number of different proteins and coordinate complex cellular processes. Protein

kinases typically mediate the phosphorylation of serine, threonine or tyrosine residues of

substrate proteins, and this has been the basis for further classification into specific

families. The family of protein-tyrosine kinases is found only in Metazoans and consists

of two main groups; receptor tyrosine kinases and cytoplasmic tyrosine kinases. There

are ten distinct subfamilies of cytoplasmic tyrosine kinases that are grouped according to

common structural domains (Figure 1.1). The Fps/Fes and Fer kinases are distinct from

the other protein-tyrosine kinases, and make up subfamily IV. Both kinases contain an

amino-terminal F-BAR (FCH-Bin/Amphiphysin/Rvs) homology motif, followed by a

predicted coiled-coil motif, a central Src-homology 2 domain and a C-terminal kinase

domain.

The discovery that Fps/Fes was prominently expressed in cells of the

hematopoietic system led to investigations to determine its biological function in this

system. The activation of Fps/Fes has been observed downstream of numerous cytokine

2

Figure 1.1 Cytoplasmic protein tyrosine kinases. Protein tyrosine kinases are a large and diverse multigene family found only in Metazoans. There are thirty-two cytoplasmic protein tyrosine kinases (PTK) that have been divided into ten subfamilies. All PTKs contain a highly conserved catalytic kinase domain. Members of each subfamily share unique subdomain motifs that distinguish them from other tyrosine kinases. The following domains are found in one or more subfamily: SH2 (Src homology 2), SH3, PH (pleckstrin homology), F-BAR (FCH-Bin/Amphiphysin/Rvs), FCH (Fps/Fer-Cdc42-interacting protein 4 homology), CC (coiled-coil), FAT (focal adhesion targeting), DBD (DNA binding domain), ABD (F-actin binding domain), Cdc42 BD, P (proline rich region), Myr (myristoylation sequence). Information for this figure was obtained from www.kinase.com.

3

receptors and it is thought to function in a variety of different processes including cell

differentiation and survival, cell adhesion and cell-cell signaling. Despite this compilation

of work, the clear physiological functions of Fps/Fes remained enigmatic. Analysis of

transgenic knock-in, knock-out and over-expression mouse models has identified Fps/Fes

as a regulator of inflammation, innate immunity, hematopoiesis, angiogenesis and

vasculogenesis. Though once thought to be restricted to the hematopoietic system,

Fps/Fes expression is now recognized to be more widespread. This is supported by the

identification of Fps/Fes in the mouse mammary gland and its ability to function as a

tumor suppressor in this tissue. However, a clear biochemical function for Fps/Fes has

yet to be established. The purpose of my research project was to characterize the

physiological and biochemical functions of Fps/Fes in the mouse mammary gland.

1.2 The Fps and Fer tyrosine kinases

1.2.1 The discovery of fps/fes as a component of retroviral oncogenes

fps and fes were first isolated as the oncogenic components of tumor-causing

avian and feline retroviruses, respectively 1-3. This was the basis for adopting the names,

fps (Fujinami poultry sarcoma) and fes (feline sarcoma). Through sequence comparisons

it was shown that these different oncogenes actually corresponded to the same gene 4. All

known viral fps/fes (hereafter referred to as fps) oncogenes encode chimeric proteins that

consist of N-terminal retroviral Gag sequences fused to a portion of, and sometimes all

of, the Fps sequence (Figure 1.2) 5-8. The Gag protein has two important effects on the

function and activity of the Fps oncoprotein. It possesses a plasma membrane-targeting

4

Figure 1.2 Comparison of cellular Fps and Fer kinases and viral Fps oncoproteins. The Fps and Fer kinases are composed of four distinct structural domains. Beginning at the N-terminus is the F-BAR domain which contains the FCH and CC1 motifs. This is followed by the CC2 motif and the SH2 and kinase domains. In contrast, all viral proteins contain a Gag sequence at their N-termini which is fused to a portion or all of the cellular Fps sequence. Poultry Research center strain (PRC), Gardner-Arnstein virus (GA), Fujinami sarcoma virus (FSV).

5

and -binding domain 9. It also confers unregulated tyrosine kinase activity on the

protein,possibly by interfering with negative-activity regulatory elements found in the

non-catalytic domains of Fps 5. Transduction of fibroblasts with the Fujinami sarcoma

virus (FSV) resulted in increased cellular phosphotyrosine content and morphological

transformation 10. Expression of Gag-Fps in lung fibroblast cells promoted growth factor-

independent growth and transformation 11. The Gag-Fps protein also induced the

differentiation of hematopoietic progenitors into macrophages in the absence of

exogenous cytokines or growth factors 12, 13. Based on their increased tyrosine

phosphorylation, a number of potential substrates have been identified in cells

transformed with viral Fps. They include p120RasGAP 14 and two associated proteins,

p62Dok and p190RhoGAP 15, 16, PI3K 17, Shc 18, connexin 43 19, BCR 20 and Stat3 21, 22.

The generation of transgenic mice in which viral Fps was expressed under the

transcriptional control of the human beta-globin promoter caused both benign and

malignant tumors in a variety of different tissues of lymphoid and mesenchymal origin.

Lymphomas, thymomas, fibrosarcomas, angiosarcomas, hemangiomas, and

neurofibrosarcomas were each observed with varying penetrances and latencies 23.

Interestingly, some tissues that expressed the v-fps oncogene, such as heart, brain, lung,

and testes, developed no malignant tumors. Thus, viral Fps possesses a range of

oncogenic activities, but its penetrance is somewhat restricted and additional genetic or

epigenetic events are likely required to promote full transformation.

6

1.2.2 The fps and fer protein-tyrosine kinase-encoding proto-oncogenes

Studies on viral Fps soon led to the discovery that parts of the oncogenic sequence

were homologous to regions of cellular DNA which at the time had no well-defined role

in normal cells 1, 4. Using the viral fps sequence as a probe, its cellular homolog was first

isolated from poultry cells 24. Since then, the human 25, feline 7 and mouse genes 26, 27

have been cloned. The human fps proto-oncogene consists of 19 exons, with exon-1

being non-coding, and spans approximately 13 kb of genomic DNA (Figure 1.3).

Transgenic mice generated with this genomic fragment demonstrated that it was

sufficient to generate the appropriate tissue-specific pattern of human fps expression that

was independent of the integration site but dependent on transgene copy number 28. Thus,

the full fps locus control region (LCR) is likely contained in this fragment. More

specifically, a minimal 2.5 kb LCR element within intron 1 to intron 3 (nucleotides +28

to +2523 relative to the transcriptional start site) also conferred tissue-specific fps

expression 29.

Much of the work elucidating the transcriptional regulation of fps expression has

focused on its activation in the myeloid lineage. With the use of deletion constructs,

combined with DNase I footprinting assays and electrophoretic mobility shift analysis, a

myeloid-specific promoter region was identified within the 151 bp directly upstream of

the transcriptional start site of the human fps gene 30. Two transcription factors, Sp1 and

PU.1, were shown to bind to, and trans-activate, the fps promoter in myeloid cells 30, 31.

PU.1, a member of the Ets transcription factor family, appears to be expressed

7

Figure 1.3 Organization of the fps and fer loci. (A) The complete fps locus is thought to be contained within 13 kb of genomic sequence. It consists of 19 exons, of which the first exon is non-coding. The fps-null mouse model was generated by replacing exons 15 through 18 with the PGKneo selection marker. This strategy resulted in the deletion of the kinase subdomains IV to X and predicted to result in an unstable truncated Fps protein. The exon organization of the fer locus is almost identical to that of fps, with the exception of an additional 5’ non-coding exon, the presence of an internal testis-specific promoter that regulates the expression of a ferT isoform. T1, T2 and T3 represent the first three exons of ferT. In contrast to fps, the fer locus is at least 500 kb, with much larger introns, as indicated by the open line. (B) The colour-coded exons corresponding to the different structural domains are: F-BAR (blue, green, yellow), CC2 (purple), SH2 (ice blue), and Kinase (black). Adapted from 5.

8

exclusively in cells of the hematopoietic system 32. It is critical to the development of

hematopoietic cells through the activation of many regulatory genes 33. Sp1 is a zinc

finger transcription factor that is ubiquitously expressed and has been linked to the

regulation of several hundred genes 34. Its presence outside the hematopoietic system

suggests it may be involved in activating fps expression in other types of cells where fps

has been detected. But at the present time very little is known about the transcriptional

regulation of fps, other than in myeloid cells.

Analysis of Fps expression with antisera raised against viral Fps identified two

cross-reactive cellular proteins of 92 and 94 kDa in size. Results of peptide mapping

analysis identified the 92 kDa protein as Fps whereas the 94 kDa protein was a related

but distinct protein 35, 36. Sequence comparison of human and rat cDNAs encoding this 94

kDa protein confirmed its similarities to Fps and led to its designation as the Fer (Fes-

related) tyrosine kinase 37, 38. The genomic organization of the fer gene has been

determined and the exon/intron structure is very similar to that of fps (Figure 1.3). There

are 20 exons in the fer gene with the first two exons being non-coding sequences. The fer

locus is much more expansive than the fps locus, as it spans an estimated 500 kb of

genomic sequence 5. In addition, the fer gene has an internal promoter that results in a 51

kDa testis-specific isoform of Fer, FerT 39 (Figure 1.3).

Homologs of Fer have been identified in several species ranging from mammals

to the primitive marine sponge 37, 38, 40, 41. A single fer-like gene has been identified in

such species as Drosophila melanogaster, Saccharomyces cerevisiae, Schistosoma

mansoni and Sycon raphanus 41-44. Surprisingly, analysis of genomic sequence from

9

Caenorhabditis elegans suggests that as many as 42 fer homologs may exist in this

species 40. However, these sequences encode proteins consisting of only SH2 and kinase

domains and could be evolutionary precursors to any SH2-containing PTK. Taken

together, these results suggest that fer probably evolved prior to fps from a progenitor

gene that encoded a SH2-domain-containing tyrosine kinase, which subsequently

acquired additional sequences encoding the F-BAR and CC domains 5. A progenitor of

the fps gene may have then arisen from a duplication of fer, which would have evolved

into the tissue-specific fps gene presently found in higher mammals.

1.2.3 Expression patterns and subcellular localization of Fps and Fer tyrosine kinases

Although Fps and Fer are closely related in terms of their overall sequence and

functional domain structures, there are obvious differences in their expression patterns.

While the expression of fer is thought to be ubiquitous 37, 38, fps expression is more

restricted. In fact, its expression was once thought to be limited to cells of the

hematopoietic system, specifically to cells of the myeloid lineage 36, 45. However, more

sophisticated cell isolation methods and improved analytical techniques have provided

the means to demonstrate the expression of fps in a number of both progenitor and mature

cells within the hematopoietic system. These include the pluripotent hematopoietic stem

cell 46, mature B and T cells, multiple progenitor cell types within the myeloid lineage

(Greer, unpublished), mature erythrocytes, megakaryocytes, platelets, neutrophils,

macrophage and mast cells 47-49 (Greer, unpublished). Additional studies have revealed

fps expression in embryonic and fetal mouse tissue from all three germ layers, but at later

10

stages of development, expression became more restricted and eventually was found only

in granulomonocytic cells 50. Transgenic mice expressing an activated form of Fps

developed vascular tumors. This led to the discovery that fps is normally expressed in

human and mouse vascular endothelial cells 51. An extensive search to determine the

scope of fps expression in both embryonic and adult tissues identified a number of

interesting fps-expressing cell types 52. Within the embryo, highest levels were detected

in endothelial cell precursors, or angioblasts, of early yolk sac blood islands, skeletal

muscle, epidermis, heart endocardium, chondrocytes, neuronal cells and lung epithelium

52. In adult tissue, in addition to the previously reported endothelial cell expression, fps

was observed within neuronal cells of the brain, epithelial cells of the choroid plexus and

in Purkinje cells. fps was also detected in glandular epithelium of the uterus 52 and

colonic crypt cells 53-55. Based on the dynamic membrane activity associated with these

types of cells, it was suggested that Fps might be involved in processes such as secretion,

endocytosis, transcytosis and potocytosis 52.

Fps and Fer were originally described as cytoplasmic tyrosine kinases, where they

have known functions 56, 57. However, Fps and Fer have been reported to localize to the

nucleus as well 57-60. A putative nuclear localization signal has been identified in the

kinase domain of Fer 58. No nuclear functions have been ascribed to Fps yet but Fer may

regulate cell cycle progression and/or transcription 58, 61. Myeloid cells treated with

different differentiation-inducing agents resulted in Fps oligomerization and

accumulation in the nucleus 59. Similarly, Fer has been shown to be cytoplasmic in

quiescent cells, and then enter the nucleus at the onset of S phase 58. In both of these

11

studies the N-terminal CC domains were found to be important for nuclear translocation.

CC domains are important for oligomerization 62 but their importance may suggest the

presence of an unrecognized nuclear localization signal as well. Another study using

green fluorescent protein (GFP) fusions of Fps and Fer cast doubt on the nuclear

localization of Fps and Fer 63. When Fps-GFP or Fer-GFP were expressed in Cos-1 cells,

both kinases were observed in the cytoplasm, and neither could be detected in the

nucleus, regardless of the cell cycle stage 63. This is in agreement with the report by

Tagliafico et al., where Fps was identified in the nucleus of differentiated myeloid cells,

but not in transfected Cos-1 cells 59; suggesting that the nuclear localization may be

dependent on the cell type and context. Nonetheless, some useful information was

garnered from the experiments using the GFP fusions of Fps and Fer in terms of their

cytoplasmic subcellular localization. Fer-GFP was found dispersed diffusely throughout

the cytoplasm whereas Fps-GFP was observed in cytoplasmic vesicles and in a

perinuclear region consistent with the Golgi apparatus. Additional proof for the presence

of Fps in these structures was provided by its co-localization with markers for: the trans

Golgi network, TGN38: the endoplasmic reticulum to the cis Golgi compartment, Rab1;

the exocytic pathway, Rab3A; and the early and late endocytic pathways, Rab5B and

Rab7, respectively 63. These data indicate that Fps may be a regulator of multiple aspects

of vesicular trafficking. Further evidence implicating Fps in these processes is its ability

to phosphorylate NSF, N-ethylmaleimide-sensitive factor, an ATPase that is involved in

the fusion of transport vesicles to their target membranes 64.

12

1.2.4 Structural components and kinase activity of the Fps and Fer tyrosine kinases

On a structural basis, Fps and Fer are unique amongst the members of

cytoplasmic protein-tyrosine kinase family, and have been classified accordingly into

subfamily IV 65. Depending on the species, there are slight variations in the size of Fps

and Fer. Murine Fps is composed of approximately 820 amino acids whereas Fer is

slightly longer, containing approximately 823 amino acids. Both Fps and Fer have four

recognized structural domains. Beginning at the N-terminus is the FCH-

Bin/Amphiphysin/Rvs (F-BAR) domain which includes the Fps/Fer/Cdc42-interacting

protein 4 (CIP4) homology (FCH) domain and first predicted coiled-coil (CC) motif. This

is followed by an additional predicted CC motif, a central Src homology 2 (SH2) domain

and a C-teminal kinase domain (Figure 1.2). Among the members of the cytoplasmic

tyrosine kinase family, the defining feature is of course the kinase domain. The SH2

domain present in Fps and Fer is also found in 6 of the remaining 9 subfamilies,

suggestive of a common evolutionary origin and important biological function 66. The F-

BAR domain is the unique feature of Fps and Fer which distinguishes them from other

kinases.

The F-BAR domain, also referred to as the extended FCH (EFC) domain, is

predicted to span approximately 300 amino acids within the Fps and Fer amino (N)-

termini 67. Within the F-BAR domain, the FCH domain is composed of about 90 amino

acids in the extreme N-terminus, and was initially named based on sequence homology

between Fer and CIP4 68, 69. Since then, the F-BAR domain has been identified in a

13

number of proteins that regulate cytoskeletal rearrangements, vesicular transport and

endocytosis 68-71. In fact, based on sequence homology nearly 100 proteins have been

identified that are predicted to contain this domain. The high conservation of this protein

family from yeast to mammals is indicative of a functional importance 72. F-BAR

domain-containing proteins are part of the larger BAR domain superfamily that includes

N-BAR and I-BAR domain-containing proteins 73. In general, BAR domain-containing

proteins are classified as regulators of membrane processes 74 and have the ability to

tubulate and shape membranes 75. Proteins containing F-BAR domains are further

classified into three general groups: the Fps/Fer subfamily of cytoplasmic protein-

tyrosine kinases; a subset of the RhoGAP proteins; and a group of membrane- and

cytoskeletal-associated scaffold/adaptor proteins 5, 76.

One of the founding FCH domain-containing proteins, CIP4, was originally

identified in a yeast two-hybrid system as an interacting partner with activated Cdc42, a

Rho-family GTPase 68 that is important in filopodia formation, membrane ruffling and

directional migration 77. In terms of cytoskeletal reorganization CIP4 can function as an

adaptor protein that interacts with proteins that regulate the cytoskeleton. For example, in

addition to Cdc42, CIP4 has been shown to interact with RICH-1, a GTPase-activating

protein for Cdc42 and Rac1 78, and the Wiskott-Aldrich syndrome protein (WASP) 71 that

activates the Arp2/3 complex to promote actin polymerization. The FCH domain of CIP4

is able to bind microtubules and, in doing so, can mediate the association of WASP with

microtubules 71. Loss of either the CIP4 FCH or WASP-binding domain in human

macrophages alters the microtubule system and results in the failure of actin-rich

14

podosome formation 79. The F-BAR domain can bind to phosphatidylinositol 4,5-

bisphosphate (PI(4,5)P2). This feature of CIP4 is critical to its ability to bind membrane

surfaces and promote membrane tubulation 80, 81. Consistent with its membrane binding

ability, knockdown of CIP4 by RNA interference resulted in reduced endocytosis, based

on the reduced internalization of fluorescently labeled epidermal growth factor (EGF) 67.

Recently the crystal structures of the F-BAR domains of CIP4 and a closely

related protein, formin binding protein 17 (FBP17), have been determined and have

provided some interesting insight into the mechanism by which these proteins promote

membrane curvature associated with endocytosis and tubulation 80. The structures

displayed a “gently curved helical-bundle dimer” which formed filaments through end-to-

end interactions 80. This is predicted to result in the formation of a circular or spiral

structure as additional F-BAR dimers are incorporated into the filament 80. So, as

membrane invagination occurs during endocytosis the F-BAR domain-containing

proteins bind the curved surface, polymerize and wrap around the growing structure to

support and drive invagination further. The so-called “banana-like shape” of the F-BAR

domain has similarly been identified in the BAR-domain dimer which also binds to the

membrane with its concave surface 81. Interestingly, the shape of the BAR domain has a

sharper curvature and a greater arc depth, which directly relates to the smaller diameter of

tubules generated by BAR domains 80. Thus, whereas filaments of F-BAR domains are

predicted to be involved in steps of endocytosis when the membrane curvature is less

severe, it is possible that the BAR domain proteins become involved as the membrane

curvature becomes greater prior to scission of the vesicle 80.

15

The F-BAR domains of Fps and Fer share some, but not all, of the characteristics

associated with those in CIP4 and FBP17. By “threading” the sequence of the F-BAR

domain of Fps onto the known structure of the CIP4 F-BAR domain, it is predicted that

the five α-helices found in CIP4 are also found in Fps, and Fer as well (Andrew Craig,

unpublished). This is particularly intriguing as Fps has long been thought to be involved

in vesicular trafficking 63 and the three-dimensional similarities with an established

regulator of endocytosis provide support for this theory. Additionally, Fps can interact

with tubulin and co-localize with microtubules that involves the F-BAR domain 82, 83.

Despite these parallels, there are distinct differences between the F-BAR domain

proteins. The positions of at least six basic amino acids are highly conserved within many

of the F-BAR domains and are essential for liposome binding and tubulation 67, 74, 81. In

contrast, in Fps and Fer only three of the six residues are conserved 67 (Andrew Craig,

unpublished). The functional significance of this is not known, but it is likely a feature

that distinguishes the Fps and Fer F-BAR domains from other F-BAR domains. In this

regard, extensive membrane tubulation was generated upon expression of the CIP4 F-

BAR domain in Cos-7 cells whereas the Fer F-BAR domain generated minimal

tubulation 67, suggesting that the F-BAR domain of Fps and Fer may not directly regulate

membrane curvature. Similar to this result is the limited ability of the Fer F-BAR domain

to induce tubulation of protein-free brain lipid liposomes 67. These results indicate that

the F-BAR domain of Fer can directly interact with lipids. More specifically, it has been

shown to bind to PI(4,5)P2 67, and Fer constructs that lacked the FCH fragment were no

longer able to bind any form of phospholipid species (Michelle Scott, unpublished).

16

Similarly, the F-BAR domain of Fps also bound PI(4,5)P2 but bound PI3P, PI4P and

PI5P to an even greater extent (Andrew Craig, unpublished). Thus, although Fps and Fer

do not share all the characteristics of other F-BAR domain-containing proteins, the

underlying theme is that they likely possess the ability to bind lipids within the

membrane. In doing so, this may allow them to interact with and phosphorylate other

proteins involved in regulating membrane processes.

Among the cytoplasmic tyrosine kinases, only Fps and Fer are known to contain

CC domains. Typically a CC domain contains a characteristic seven amino acid repeat

sequence that forms a multi-α-helical structure to mediate oligomerization 84. Depending

on the algorithm used to analyze the amino acid sequence, the amino-termini of Fps and

Fer are predicted to contain either two or three CC domains 62, 85. However, these

predictions were performed prior to the identification of the F-BAR domain. The CC

domain in the F-BAR domain-containing proteins, corresponding to the CC1 domain in

Fps and Fer, is predicted to form an α-helical structure and is involved in protein-protein

interaction but the recent studies showing that it is actually part of the F-BAR domain

cast some doubt on its structure and function as a “classical” CC domain 67, 80. Despite

this, the N-terminal regions of Fps and Fer have been shown to be important for

oligomerization 62, 85. Fps is thought to exist in several different forms, ranging from

monomers to pentamers and possibly higher order oligomers 85, 86, whereas Fer is

constitutively found as a trimer 62. Oligomerization does not regulate the activity of Fer

but it is likely a prerequisite for activation of Fps 86. Despite the homology between Fps

and Fer there is no evidence of heterotypic interactions 62. Interestingly, a lysine to

17

proline mis-sense mutation within the first or second CC domains of Fer abolishes its

oligomerization ability, suggesting that both domains are involved in this process 62. The

same mutation in the first CC domain of Fps actually caused increased kinase activity 86.

It is unlikely that this mutation hindered oligomerization because it resulted in increased

activity. However, it is consistent with a role for the first CC in the negative regulation of

Fps kinase activity.

Through a series of experiments involving mutation, deletion and over-expression

of the CC regions, a working model has been developed by Smithgall et al., that explains

how the CC domains of Fps regulate tyrosine kinase activity 87. They suggest that when

Fps is inactive it is found as a monomer, and that oligomerization is required for kinase

activity. Inactive Fps is thought to be maintained as a monomer through intramolecular

cis-interactions where CC1 and CC2 interact with one another and inhibit trans-

interactions. Fps activation would then involve its recruitment to a specific location, eg.

ligand-bound receptor, followed by disruption of the intramolecular CC interactions and

formation of oligomers through intermolecular CC interactions 87. The oligomerized Fps

could then become auto-phosphorylated in trans and phosphorylate downstream

signaling molecules.

The SH2 domain is a non-catalytic domain of approximately 100 residues that is

highly conserved throughout evolution 66, 88. It is found in a wide variety of signaling

molecules and functions to mediate protein-protein interaction by binding to

phosphotyrosine residues 89. The SH2 domain was first described in v-Fps and was

known to be important to kinase activity and transformation 66, 90. Within Fps it is located

18

immediately N-terminal to the kinase domain. In addition to its role in binding putative

phosphotyrosine-containing substrates the SH2 domain likely regulates the kinase

activity of Fps 91. That a Fps mutant lacking the SH2 domain displayed a marked

reduction in auto-phosphorylation and substrate phosphorylation activity 91 supports this

theory.

Screening of a phosphopeptide library using the Fps SH2 domain as an affinity

matrix has identified pYEXV/I (in which pY represents phosphotyrosine; E is glutamate;

X is any amino acid; V is valine and I is isoleucine) as the consensus Fps SH2 domain

binding sequence 92. This sequence has been found in a large number of proteins,

including Abl, BCR, CD3ε, CD72, ezrin, FAK, FcγRI, γ-adaptin, Hck, LAR-PTP, Lyn,

SHP-1, Tec and 3BP2A92. However, despite the presence of the consensus sequence in

these proteins there is direct evidence to suggest that only a few are authentic interacting

partners of Fps.

To determine the consensus substrate sequence recognized by Fps, purified Fps

protein was incubated with a peptide library and the interacting peptides were then

separated and sequenced 93. The results showed that the optimal substrate sequence was

very similar to the consensus recognition sequence of the SH2 domain, suggesting that

potential interacting proteins are also good candidates for phosphorylation.

There are at least two tyrosine phosphorylation sites in the catalytic domain of

Fps 91, 94. The first is found in the activation loop at Y713 in Fps, corresponding to Y715

in Fer 62, 95. It is highly conserved among kinases and is essential to kinase activity as

suggested by the reduced activity upon mutation 94, 95. The second site is found at Y811 in

19

Fps 96. This site does not regulate kinase activity but is part of a sequence that may act as

a docking site for other proteins with SH2 domains 92, 96. Both sites can be

phosphorylated by Fps itself as a result of trans autophosphorylation 96.

Fps can also be phosphorylated by other tyrosine kinases. For example,

stimulation of macrophage expressing a catalytically inactive Fps with granulocyte-

macrophage-colony stimulating factor (GM-CSF) results in tyrosine phosphorylation of

Fps (Greer, unpublished). Similarly, FcεRI receptor activation in mast cells leads to

phosphorylation of both kinase-dead Fps and Fer that is highly dependent on the Src-

family kinase, Lyn 97. The kinase activity of Fps is thought to be tightly regulated through

intramolecular interactions with the CC1 and SH2 domains 86, 98. The kinase activity of

Fps is also relatively weak when compared to that of Fer. This has been demonstrated by

over-expression studies in cell culture systems where the phosphotyrosine level of Fer is

much higher than Fps (Greer, unpublished). Similar results have been obtained when

each protein is subjected to in vitro kinase assays (Greer, unpublished).

1.2.5 Transgenic mouse models of the Fps tyrosine kinase

Over the past two decades several mouse models have been generated with the

purpose of developing a better understanding of the role(s) of Fps during both normal and

pathological conditions. The first mouse model of Fps was designed to determine the

consequences of Fps over-expression. These mice were generated from a single cell

embryo micro-injected with the 13 kb fragment of human genomic DNA encoding the fps

gene 28. The resultant mouse line contained approximately 15 to 20 copies of the human

20

fps allele and displayed the same tissue-specific expression pattern as the endogenous

mouse locus 28, 52. Despite the increase in Fps expression no phenotypic changes were

evident. But the study was able to demonstrate that the 13 kb fragment of DNA contained

the entire fps locus.

The lack of a phenotype associated with Fps over-expression was probably due to

the normally tightly regulated kinase activity of Fps. This led the same group of

investigators to insert a Src myristoylation sequence into the 5’ end of the 13 kb human

fps genomic clone that resulted in the expression of a mutant hyperactive Fps 51.

Expression of this Fps protein in Rat-2 fibroblast cells resulted in their transformation.

Thus, it was expected that mice expressing this construct would provide evidence that

Fps was involved in tumorigenesis. Surprisingly, the most obvious characteristic of these

mice was widespread hypervascularity which developed into multifocal hemangiomas.

Prior to this study, Fps was not known to be expressed in the vascular system. But,

through the use of RNase protection and in situ hybridization analyses, fps transcripts

were detected in both normal and tumor vasculature 51, 52. From these results it was

suggested that Fps may regulate angiogenesis and/or vasculogenesis. Additional studies

showed that there was an approximately two fold increase in vascularity due to a greater

number of secondary vessels in these mice and that Fps was a component of the VEGF

and PDGF signaling pathways 99. Endothelial cells from yolk sacs expressing the

activated Fps were readily established in culture further implicating Fps in endothelial

mitogenic pathways 100. These mice have been used to provide evidence showing Fps has

a role in lineage determination of hematopoietic progenitors, immature erythroid

21

precursors and the more committed granulomonocytic progenitors 48. The results derived

from the mutant hyperactive Fps mice have provided interesting clues as to the in vivo

function of Fps. However, the effects of myristoylation on protein function should not be

discounted as it may confer characteristics not associated with the normal Fps protein.

The generation of a third mouse model by the Greer group involved replacing a

highly conserved lysine residue to arginine at position 588 in the ATP-binding fold of the

Fps catalytic domain, which resulted in a stably expressed but kinase-inactive protein 101.

The rationale behind this mouse model was that kinase-inactive Fps should function in a

dominant-negative manner, enabling the identification of the signaling pathways and

interacting partners of Fps. With a focus on the hematopoietic system, it was shown that

in bone marrow-derived mutant macrophages the tyrosine phosphorylation of the signal

transducer and activator of transcription 3 and 5A (Stat3 and Stat5A) in response to GM-

CSF was dramatically reduced, as was the lipopolysaccharide-induced Erk1/2 activation

in the mutant macrophage 101. Despite the biochemical observations there was no real

difference in the levels of myeloid, erythroid, or B-cell precursors. Together these data

suggest that Fps has only a minor role in myelopoiesis and the immune response.

Theoretically it is possible that Fps possesses abilities independent of its kinase

activity that could not be revealed with the kinase-inactive model. With the hope of a

more dramatic phenotype, a fps-null mouse model was generated by the Greer lab that

addressed this issue. The deletion of exons 14-18 resulted in a fps-null mutation where no

detectable protein was produced. Macrophage from these mice showed normal GM-CSF,

IL-3 and IL-6-induced activation of Stat3 and Stat5A. However, the male fps-null mice

22

were significantly more sensitive to LPS challenge than wild-type males 26, suggesting

that the loss of Fps compromised the normal innate immune response.

Subsequently it was shown that the LPS treated fps-null mice exhibited elevated

pro-inflammatory cytokine secretion, namely TNF-α, which was associated with

prolonged degradation of IkappaB-α, increased phosphorylation of NFkappaB (p65

subunit) and reduced Toll-like receptor 4 internalization 102. Increased leukocyte

extravasation and migration into the peritoneal cavity were also observed 103. Together,

these results point to Fps as having a number of divergent functions that regulate the

immune response.

A second fps-null mouse model has been generated by another group of

investigators 104. Surprisingly, a much more apparent phenotype was identified in these

mice which displayed: partial embryonic lethality; dramatic lymphoid and myeloid

homeostasis defects; compromised innate immunity; increased Stat3 and Stat5 activation

in response to GM-CSF and IL-6 treatment; cardiomegaly; and ulcerated skin lesions 104.

The contradictory phenotypes found in this model compared to the one generated by the

Greer lab are not understood. However, genetic differences in the mouse strains may

have some role. Different gene targeting strategies could also be a factor. The second fps-

null model was generated by targeting the 5’ end of the gene. Within 1400 bp upstream

of the fps transcriptional start site is the furin locus, and any loss of neighboring DNA

could possibly affect furin expression. Furin is a proprotein convertase that cleaves a

variety of larger proteins into their mature active forms 105. furin knockout mice have a

23

number of severe defects that cause embryonic lethality, a phenotype observed in the

second fps-null mouse model 104, 106.

1.2.6 The Fps tyrosine kinase in the mammary gland

A study designed to identify kinases that were developmentally regulated in the mouse

mammary gland determined that fps mRNA levels increased during pregnancy and

reached a maximum during lactation 107. This was the first report showing that Fps was

expressed in the mammary gland. To determine if Fps regulated some aspect of breast

cancer, mice with either the fps-null and kinase-inactive mutations were crossed with the

mouse mammary tumor virus (MMTV) polyoma virus middle T antigen (PymT) breast

cancer mouse line 55. The PyVmT line develops multifocal mammary tumors with 100 %

penetrance in females 108. As expected, tumors developed in all mice that expressed

PyVmT. However, in the fps-null and kinase-inactive backgrounds, tumors developed at

a significantly earlier onset than their wild-type counterparts 55. These results indicate

that Fps may function as a tumor suppressor in the mammary gland. This was very

surprising as Fps had always been thought of as having oncogenic potential.

Another study that examined the catalytic domain sequences of 89 different

tyrosine kinases in human colorectal cancers identified four different somatic mutations

in fps 53. Subsequent investigations into the effects of these mutations showed that all

four resulted in a complete loss of in vivo kinase activity 54, 55. If these mutations could be

directly linked to the progression of colorectal cancer then it would provide addition

evidence to suggest that Fps may function as a tumor suppressor.

24

1.3 The mammary gland

The mammary gland evolved in mammals as an organ to nurture the offspring

during their postnatal development. Unlike other organs in the body the mammary gland

is unique in that much of its development involving growth, tissue organization,

differentiation and apoptosis occurs throughout adult life, with multiple rounds of

development and regression over the reproductive life of most species 109.

The mammary gland is composed of a variety of cell types that can be generally

categorized into epithelial and stromal compartments (Figure 1.4). The epithelium

consists of two cell populations, luminal epithelial cells and basal/myoepithelial cells 110.

The luminal epithelial cells are cuboidal or columnar in shape and line the ducts and

alveoli that are responsible for milk secretion. More than one layer of luminal epithelial

cells may line the primary and secondary ducts but become fewer with higher orders of

branching. Typically, the terminal ducts have only one layer of luminal cells 109. The

elongated myoepithelial cells form a continuous monolayer sheath around the ductal

epithelial cells 111. In the alveoli however, the myoepithelial cells form a basket-like

network and in between their cell processes the luminal epithelial cells are in contact with

the basement membrane 112.

The stroma consists of both a cellular and an acellular fraction. Stromal cells

found in the mammary gland include adipocytes, fibroblasts, various cell types that make

up blood and blood vessels and nerve cells 113. Acellular components include proteins

like laminin, fibronectin, type IV collagen and heparin sulfate proteoglycans that make up

the basement membrane at the epithelial-stromal boundary 114. An intact basement

25

Figure 1.4 Cellular components of the mouse mammary gland. The upper panel depicts the whole inguinal mammary gland, showing the characteristic lymph node, and epithelial ducts. The lower panel is a magnified image of a typical ductal cross section. Cuboidal luminal epithelial cells line the inner lumen and are surrounded by a basal layer of myoepithelial cells. Multiple types of cells make up the stroma which is separated from the epithelial compartment by the basement membrane.

26

membrane is important, in part, for structural support of the mammary cells. Along with

the cellular components of the stroma, it also plays an important role in mammary gland

development and function by regulating epithelial cell growth, shape, differentiation,

polarity and responsiveness to hormones 115. Stromal alterations have been shown to not

only disrupt normal mammary function but also actively contribute to tumorigenesis 113.

A reciprocal relationship exists between the epithelium and stroma such that each

influences and depends on the other 116. Thus, some compounds, such as certain growth

factors, hormones and enzymes that are produced by one compartment, can specifically

act on the other compartment to influence its function.

1.3.1 Mammary gland development

Much of the knowledge regarding mammary development has been obtained from

the analysis of transgenic and knockout mice. Therefore, unless otherwise stated, the

information described below is in reference to the mouse.

1.3.2 Early mammary gland development

Mammary development is initiated during embryogenesis during which time

mammary buds form at the sites of future nipples and then begin to elongate and

penetrate into the fat pad precursor 117. By birth, a simple structure has developed that

consists of 10-20 branched epithelial ducts within the proximal end of the fat pad. Each

duct is composed of at least one layer of luminal epithelial cells surrounding a central

lumen 118. Located beneath the epithelial cells is a layer of myoepithelial cells that rests

27

on the basement membrane and at this stage forms a continuous sheath around the

epithelial duct to separate the epithelial and stromal compartments 119. The important

biochemical regulators of embryonic mammary development have not been extensively

studied, though it is well established that mammary development in the embryo is

hormone independent 120.

After birth, very little development of the mammary gland takes place until the

onset of puberty at 3 weeks of age 121. At puberty a second round of ductal branching

morphogenesis is initiated under the influence of the ovarian hormones, estrogen and

progesterone, and the pituitary-derived growth hormone (GH) 122. During this stage

multicellular bulbous-shaped terminal end buds (TEBs) form at the tips of the ducts and

penetrate further into the fat pad as the ducts elongate 123. Ductal elongation is regulated,

in part, by estrogen which acts through the estrogen receptor (ER) to stimulate

proliferation of epithelial cells in the TEBs and the surrounding stromal cells 124, 125. ERα

activation in the epithelium upregulates the transcription of the epidermal growth factor

receptor ligand, ampiregulin, which then acts via a paracrine mechanism to regulate

epithelial proliferation, TEB formation and ductal elongation 126. GH and its receptor

(GHR) are important for both TEB formation and ductal branching 127. However, the

epithelium is not a direct target of GH. Rather, GHR-positive stromal cells become

activated to produce and release insulin-like growth factor-I which then stimulates

epithelial cells 128. Extension of the ductal system is accompanied by the formation of

new ducts through the bifurcation of the TEBs. Secondary side-branches then sprout

laterally from the established ducts until the entire fat pad is penetrated with an extensive

28

ductal system. Thereafter, additional short tertiary branches sprout from along the ducts

to further fill out the ductal tree 109. Prolactin and the prolactin receptor regulate ductal

morphogenesis during the puberty stage 129, 130 . Loss of either component results in the

absence of side-branching, the persistence of TEB structures beyond the normal time

frame, and a lack of alveolar bud formation along the ductal tree 129-131. However, some

of the defects are associated with the disruption of ovarian function observed in these

mutants 132. Both estrogen and progesterone are also involved in promoting ductal side-

branching 133, 134. Interestingly, estrogen induces progesterone receptor (PR) expression in

the ductal epithelium. While each hormone alone is capable of promoting proliferation,

complete alveolar development requires the combined effects of both hormones 133.

During the post-puberty stage, side-branching and alveolar bud formation are stimulated

with each estrus cycle which has a cumulative effect on the alveolar epithelial content 135.

With sexual maturity that occurs by 10-12 weeks of age, most TEBs have reached the

edge of the fat pad, stopped dividing and regressed to form terminal ducts. By this time,

the alveolar buds have formed rudimentary structures that have the capacity to become

fully differentiated milk secreting alveoli during pregnancy and lactation 136.

1.3.3 The mammary gland during pregnancy and lactation

Much like post-puberty development, mammary growth during early pregnancy

involves proliferation of ductal branches and alveolar bud formation; however, it occurs

on a much more massive scale 119. Both the pituitary gland and ovaries are essential for

this process. It is well established that prolactin is necessary for the proliferative phase of

29

alveolar development 137. Estrogen and progesterone are also important for proper

development during pregnancy. A recent mouse model where the ERα was deleted in the

epithelium during pregnancy demonstrated that estrogen was important for tertiary

branching, along with the proliferation and full differentiation of alveolar cells 138.

Unfortunately, the signaling pathways activated by estrogen during pregnancy were not

reported. Progesterone has been shown to promote epithelial cell proliferation during

pregnancy that results in side-branching and alveologenesis 139, 140. Interestingly,

activation of PR-positive epithelial cells results in the proliferation of surrounding cells

via a paracrine mechanism 141. Wnt-4 was recently identified as a progesterone inducible

gene in epithelial cells and mice lacking Wnt-4 display defects in ductal branching

similar to PR-/- mice 142, suggesting that it is one of the paracrine effectors of

progesterone.

From mid to late pregnancy individual alveoli progressively differentiate from the

alveolar buds and occupy most of the space of the fat pad by late pregnancy 143. This is

considered the secretory differentiation phase of mammary development 144. Prolactin is

essential for full lobuloalveolar proliferation and functional differentiation. In the absence

of either prolactin or the prolactin receptor, no alveolar development takes place 131, 141,

though side branches and alveolar buds form normally. The cellular response to prolactin

stimulation involves the coordinated activation of several different signaling pathways.

Initially, the binding of prolactin to its receptor causes receptor dimerization followed by

the recruitment and activation of janus-2 kinase (JAK2) which, in turn, phosphorylates

both the receptor and Stat5 145. Of the two forms of Stat5

30

, Stat5a has a more important role in mammary development than Stat5b but both

contribute to the full function of the gland 145, 146. Once activated, Stat5 dimerizes,

translocates to the nucleus and participates in the transcriptional activation of multiple

genes involved in the alveolar differentiation process, including establishment of

epithelial polarity and cell-cell interactions, as well as those involved in milk production

147-149.

The combined use of transgenic mouse models and transcriptional profiling

studies has identified some of the key components and effectors of the prolactin signaling

pathway. Several transcription factors are either induced, including the GATA binding

protein-3, activator protein-2 gamma and E74-like factor 5 (ELF-5) 131, 150, or activated,

nuclear factor 1-C2 (NF1-C2) 151, 152, by prolactin stimulation. In turn, these factors

promote the expression of additional proteins that are required for further differentiation.

For example, NF1-C2 induces the expression of milk genes including whey acidic protein

(WAP) 151. ELF-5 has been shown to be essential for mammary development during

pregnancy and its over-expression can rescue the defects associated with the loss of the

prolactin receptor 137, suggesting that they are components of the same pathway and ELF-

5 is an important mediator of the prolactin signal.

Prolactin induces the expression of a number of secreted growth factors and

ligands, including Wnt-4, amphiregulin, receptor activator of nuclear factor-kappa B

ligand (RANKL) and IGF-2 131, 153. Each of these factors is known to regulate one or

more aspects of mammary development 153-156. Prolactin also induces several proteins

31

like keratins and collagens involved in regulating cellular structure, and specific claudin

and connexin proteins that regulate cell permeability and communication, respectively

131.

Of course, the expression of many of the milk components is regulated by

prolactin. Milk protein expression has been shown to be dependent on both prolactin

stimulation and phosphorylation of Stat5 145, 157, 158. These proteins fall into two

categories. The first group is initially induced during early- to mid-pregnancy and

includes Wdnm1, most of the caseins and WAP, to name a few 136, 159. The expression of

the second group of proteins is activated late in pregnancy or shortly after parturition, and

includes α-lactalbumin, δ-casein, Mucin-1 and PTHrP 160. The significance of the

differential timing patterns is not known but it may suggest addition pregnancy-related

functions for these “early” proteins. In addition to its role as a milk protein, α-

lactalbumin is the regulatory unit of the lactose synthetase enzyme. α-lactalbumin

deficient mice produce a thick milk absent of lactose and fail to sustain their offspring 161.

The synthesis of lactose involves the use of glucose as the primary substrate.

Consequently, to satisfy the cells need for glucose for lactose synthesis the glucose

transporter 1 (GLUT1) becomes highly upregulated late in pregnancy 162 and at

parturition. GLUT1 localizes to the cellular basal membrane and the Golgi membrane to

transport glucose into the cell and the Golgi, respectively 163.

Lipids are another important component of milk and are vital to the health of the

offspring. Much like the regulation of milk protein synthesis, a hierarchical system exists

to regulate the production and secretion of milk lipids. The expression of several enzymes

32

involved in lipid synthesis is increased late in pregnancy to accommodate the needs of the

gland during lactation 164. Akt1 has an important role in upregulating lipid synthesis and

downregulating lipid catabolic enzymes during lactation 165. Interestingly, mice that lack

Akt1 or express an activated form of Akt1 fail to lactate normally. As expected the loss

of Akt1 results in reduced milk fat and volume, whereas the activated form results in milk

with two to three times the normal fat content 165, 166. Akt1 has been shown to directly

phosphorylate some of the enzymes involved in lipid synthesis and may regulate their

activity 167. More importantly however, is its ability to upregulate and activate the

transcription factor, sterol regulatory element binding protein 1 (SREBP1), which plays

an important role in the transcriptional activation of multiple fatty acid biosynthesis genes

168, 169. How Akt1 regulates SREBP1 is not entirely known but it may be an indirect

effect. SREBP1 can be phosphorylated by glucose synthase kinase 3 (GSK-3) which

targets it for ubiquitination and degradation, and results in reduced transcriptional activity

170. GSK-3 is an Akt1 substrate, and phosphorylation by Akt1 inhibits its catalytic

activity 171. Thus, Akt1-mediated inhibition of GSK-3 reduces SREBP1

phosphorylation/degradation and activates the transcription of SREBP1-dependent fatty

acid synthesis genes 169.

By day 18 of pregnancy, the mouse mammary gland has entered the secretory

activation stage, at which time all of the protein and lipid components of the milk are

being synthesized in preparation of lactation 144. Prior to this, the presence of

progesterone suppresses milk secretion 157. It is at this stage that progesterone levels

dramatically decline to promote milk secretion and contribute to parturition. Prolactin

33

levels rise at this time and are essential for lactation 157. Late in pregnancy the gland is

characterized by increased expression of milk proteins, closure of tight junctions between

the epithelial cells and the release of lipid drops and casein micelles into the lumen 172.

The myoepithelial cells still surround the alveoli but are no longer a continuous sheath

173. This allows the epithelial cells to establish direct contact with the basement

membrane which is important for terminal differentiation and milk secretion 174. The day

before parturition, the rough endoplasmic reticulum, Golgi and secretory vesicles

accumulate in the epithelial cells and milk secretion is initiated 175, 176.

The release of milk causes the alveolar lumens to become engorged. Neonate

suckling induces; 1) a further increase in expression of most of the genes involved in

milk secretion and 2) the release of oxytocin into the blood from the hypothalamus 177, 178.

The myoepithelial cells around each alveolus contract in response to oxytocin to force the

milk out and into the ducts 179. The accumulation of milk and the cumulative effects of

the contracting myoepithelium push the milk further along the ductal system until it

reaches the nipple opening and is released 180. This process continues for approximately

three weeks until the pups are weaned.

1.3.4 Involution

Once the pups have been weaned and milk removal has stopped, the gland

undergoes involution 181. Mammary involution is thought to occur in two stages. In the

first, milk accumulation activates an initial apoptotic phase that lasts approximately 48

hours but can be reversed if suckling is re-established 182. However, continued inactivity

34

beyond this time causes the gland to enter a non-reversible phase where it is committed to

undergo widespread apoptosis and tissue remodeling that results in the regression of

much of the lobuloalveolar system.

Involution involves an up-regulation of pro-apoptotic factors and a reduction of

survival factors 183. Several genes, including Stat3 184, C/ebpδ 185, p53 186, Tgfβ3 187, and

Vitamin D3 receptor 188 are important to the activation and/or progression of the early

phase. Stat3 activation occurs very early in involution in response to the presence of

leukemia inhibitory factor (LIF) 189 and subsequently induces the expression of several

apoptotic effectors 190. The Stat3 target gene CCAAT/enhancer binding protein delta

(C/ebpδ) is a crucial mediator of pro-apoptotic gene expression. In its absence involution

is delayed, where several key pro-apoptotic genes are not activated and anti-apoptotic

genes are not repressed 191. Stat3 also inhibits the survival pathway mediated by Akt by

inducing the expression of two smaller isoforms, p50α and p55α, of the p85α regulatory

subunit of PI(3)K. These two isoforms are thought to block PI(3)K-mediated

phosphorylation of Akt and be responsible for downregulating survival activated by the

Akt pathway 192.

Epithelial cells that are beginning to undergo apoptosis are shed into the lumen.

This effect requires detachment from the surrounding epithelium and underlying

substratum 193. Disruption of E-cadherin-based adherens junctions is a key step in the loss

of cell-cell adhesion, and is known to precede apoptosis during involution 194. The

cysteine-aspartic acid protease, caspase-3, also becomes activated shortly after weaning,

but only in cells that have been shed into the lumen 195.

35

Involution has been described as resembling a wound healing process but lacking

the associated inflammatory response 196, 197. Increased neutrophil numbers can be seen

two days after pup weaning and continue to increase in number as more cells die,

followed by macrophage infiltration 196. Neutrophil-attracting chemokines have been

detected as well 197. The main function of these cells in the mammary gland may be the

removal of apoptotic cells and debris through phagocytosis 193.

After the second day the gland enters the irreversible phase of involution in which

the majority of the epithelium is removed and the tissue is remodeled 198. At this time the

epithelial cell layer displays clear signs of apoptosis, and caspase-3 activation is more

widespread 199. By day four the alveoli have collapsed into clusters of apoptotic cells and

the alveolar lumina are no longer visable. The epithelium becomes very disorganized and

continues to decrease as adipocytes begin to refill with lipid and increase in size. By day

six all the alveoli are destroyed, and the stroma and basement membrane are being

reorganized 200. This is reflected by the expression of matrix metalloproteinases 11 and

12 and carboxypeptidases E, X1 and A3 during this stage 193. The majority of epithelial

cell death has occurred by this time and will be cleared either by neighboring epithelial

cells or phagocytic macrophages 201, 202, which may require up to day 14 to be completed.

Although the majority of the epithelial cells are removed during involution the primary

ducts are maintained, even though they are known to undergo some structural

modifications 109. The process of elimination, reorganization and remodeling continues,

and after three weeks of involution the gland resembles that of a pre-pregnant mature

mouse.

36

1.4 Milk

The primary function of the mammary gland is the production and secretion of

milk. Although all mammals produce milk the complexity and composition of the milk

from each species is a reflection of the nutritional requirements of the offspring. Milk has

three main components: proteins, lipids and lactose, but also includes minerals, vitamins,

enzymes and growth factors. The mechanisms that produce the different components are

tightly regulated and activated only during pregnancy and/or lactation to produce

adequate amounts of milk 160.

1.4.1 The secretion of milk components

The secretion of milk is a complex and highly regulated process. Despite its

importance to the survival of all mammalian neonates little is known about much of the

biochemical and molecular aspects of milk secretion. Transcellular routes are

predominantly involved in this process and can be divided into four different secretory

pathways. Depending on the nature of the milk components they can be generated either

endogenously by the alveolar epithelial cells or imported into the cell from the blood

serum before being released into the lumen 203. Of the transcellular routes, Fps is most

likely to regulate some aspect of only the exocytotic and lipid secretion pathways (Figure

1.5). Because of this and due to space limitations the transcytotic and membrane transport

pathways will not be discussed.

37

Figure 1.5 Mammary epithelial cell exocytotic and lipid secretion pathways that contribute to the production of milk during lactation. Proteins originating in the rough endoplasmic reticulum (RER) and other molecules are transported into or synthesized in the Golgi apparatus (GA) where they are sorted, modified and packaged into secretory vesicles (SV). Vesicles are transported to the apical membrane where they are released into the alveolar lumen via the exocytotic pathway. Lipids are produced and released into the lumen by the lipid secretion pathway. Lipid synthesis occurs in the membrane of the smooth endoplasmic reticulum (SER). Groups of lipid molecules bud into the cytoplasmic compartment to form microlipid droplets (MLD). Droplets can grow in size by fusing with each other to form larger cytoplasmic lipid droplets (CLD). At the apical surface droplets are enveloped by plasma membrane, giving rise to milk lipid globules (MLG). It is believed that the net loss of membrane associated with the secretion of milk lipid globules is replenished when the secretory vesicle membrane becomes integrated into the plasma membrane.

38

1.4.2 The exocytotic and lipid secretion pathways

There are two pathways for the secretion of endogenously generated substances;

the exocytotic pathway and the lipid secretion pathway. The exocytotic pathway is the

primary route used by the alveolar cells for the secretion of proteins, water, lactose,

oligosaccharides, phosphate, citrate and calcium 203. Similar to that in other types of cells,

these substances are synthesized and/or packaged into secretory vesicles that are destined

for the apical region of the cell 176. Initially immature vesicles are generated by the Golgi

where they undergo processing and remodeling to mature secretory vesicles 204. The

maturation process involves both the fusion of immature vesicles, along with the sorting

and concentration of substances. The effect of this is to reduce the amount of membrane

and remove mis-sorted molecules 205.

Mature vesicles, containing various milk components, are transported to the apical

plasma membrane where the two membranes fuse and the vesicular contents are released

into the lumen. Efficient vesicular transport is dependent on intact microtubule and

microfilament systems 206-208. This is supported by the observations that disruption of

either microtubules or microfilaments blocks milk secretion 209-211. The proteins involved

in the fusion of the vesicle and apical membranes have not been studied in the mammary

gland but are thought to be similar to, or the same as, those in other systems where

membrane fusion involves the interactions of three soluble N-ethyl-maleimide-sensitive

factor (NSF)-attachment protein receptors (SNAREs) 212. The vesicle-specific SNAREs

(v-SNAREs) are found on the vesicle surface whereas target SNAREs (t-SNAREs), eg.

SNAP-25 and syntaxin, are located on the apical, or target, membrane 212, 213. After a

39

vesicle is transported to the apical membrane it becomes tethered to the membrane inner

surface at specific sites called porosomes through the association of the v-SNARE with

the t-SNAREs 214. The SNAREs on opposing membranes interact in a circular array to

form a ring complex and bring the membranes close together 215. Normally the negative

charge on each membrane generates a repulsive force; however the Ca2+ present between

the two layers creates a bridge between the phospholipid head groups 216. This leads to

the destabilization of the two lipid bilayers within the SNARE ring complex which

ultimately results in lipid mixing and membrane fusion. Following the generation of a

pore, the vesicle contents are released 217. Once secretion is completed, the vesicle may

become part of the apical membrane or detach from the membrane and be recycled to the

Golgi for future rounds of secretion 218.

In addition to the abundant milk proteins, lipids are a major source of energy for

newborns of all mammals. Milk lipids are composed primarily of triacylglycerides and, to

a lesser extent, phospholipids. During lactation, epithelial cells have highly expanded

capabilities for lipid synthesis, storage and secretion 176. The triacylglycerides are

synthesized in the smooth endoplasmic reticulum (SER) from fatty acids and glycerol by

membrane-bound fatty acyl transferases 175. Where in the membrane the newly

synthesized lipids aggregate into nascent droplets is still unclear. However, at some stage,

microlipid droplets (less than 0.5 µm in diameter) are coated with protein and polar lipids

from the SER membrane and released into the cytoplasm 219. The overall protein

composition of the lipid droplets is distinct from that of the SER, though a limited set of

proteins are shared 219. Similarly, the surface lipids are different from those of the SER

40

membrane 220. These observations provide evidence of a regulated sorting process that

targets specific proteins and lipids to the surface of the lipid droplet, which are likely

important to its formation and function.

After release from the SER, droplets can undergo a series of fusions to form larger

cytoplasmic lipid droplets during their transport to the apical region of the cell 175. These

droplets can become relatively large before secretion, but others are secreted virtually

unchanged in size from the time of their initial formation 175. While the lipid droplets are

thought to originate from the SER, they are transported to the apical membrane prior to

secretion into the lumen 221. This would suggest that specific trafficking and docking

signals are located on the lipid droplets. A proteomic analysis showed that components of

microtubule motors and cytoskeletal-interacting proteins were present on lipid droplets

and milk fat droplets from lactating mice 219. Specifically, dynein intermediate chain,

motor protein, gelsolin and gephyrin were identified. Because the dynein motor is

thought to transport cargo towards the microtubule minus-end, these authors suggested

that mammary epithelial microtubules may be oriented with their minus-ends toward the

cell periphery 219. Pancreatic acinar cells have been shown to possess a similar system to

transport secretory vesicles to their apical membrane 222.

Once at the apical surface, the lipid droplets become enveloped by the cell

membrane as they are secreted into the alveolar lumen 175. This process is essential to the

quality of the milk because in its absence lipid droplets would coalesce into large

aggregates and compromise milk expulsion. Secretion involves the coupling of the lipid

droplet to the apical plasma membrane through the formation of a multi-protein complex.

41

Three proteins have been identified in this complex; the lipid droplet surface protein,

adipophilin (ADPH); the cytoplasmic enzyme, xanthine oxidoreductase (XOR); and the

apical transmembrane protein, butyrophilin (Btn) 176, 223. Knockout mouse models have

shown that loss of either XOR or Btn causes lactational defects. In each system there is

an accumulation of cytoplasmic lipid droplets due to reduced lipid secretion 224, 225. The

mechanisms that control the release of lipid droplets into the lumen have not been studied

in depth. There is evidence that the trimeric G protein β is involved in exocytosis in

pancreatic cells 226. This protein has been identified in membrane enveloped milk fat

droplets but its function is not known 219.

1.5 Cell adhesion

1.5.1 Cadherin-mediated cell adhesion

Most cells in multicellular organisms interact preferentially with cells of their

own type, enabling them to be arranged into distinct cohesive groups with related

structures and functions. The establishment and maintenance of these groups of cells into

tissues is regulated in part by cell-cell adhesion systems 227. The ability of cells to adhere

to one another provides a mechanism to ensure that adjacent cells interact and function

appropriately. In part, cell adhesion is mediated by transmembrane proteins that bind to

identical proteins on the surface of neighboring cells 228. Members of the classical

cadherin protein family are the best understood mediators of intercellular adhesion.

Classical cadherins are single-pass membrane spanning glycoproteins that usually consist

of five highly homologous extracellular (EC) domains that are each approximately 115

42

amino acids in length 227 and a shorter cytoplasmic tail that is indirectly linked to the

cytoskeleton. Between the EC domains are located calcium binding sites 229, 230. The

binding of calcium to cadherin causes a conformational change in its shape which is

required for the interaction between cadherin molecules and for cell-cell adhesion 231.

E-cadherin is expressed exclusively in epithelial cells and its name was derived

from its expression in this cell type 232. It has a fundamental role in epithelial cell-cell

adhesion. E-cadherin molecules can form lateral homodimers within an epithelial cell or

adhesive homodimers where the extracellular domains interact between neighboring cells

233. The significance of the lateral dimers is not understood but the adhesive dimers are

required for cell to cell adhesion. Each adhesive interaction is considered to be relatively

weak but the adhesive contacts become strengthened when the E-cadherin complexes

cluster in specific membrane regions to form adherens junctions (AJ) 234.

Adhesion is thought to be strengthened further by the direct or indirect interaction

of the E-cadherin cytoplasmic domain with a network of signaling molecules, including

α-catenin, β-catenin and p120-catenin, which stabilize cell-cell adhesion and provide a

linkage to the actin cytoskeleton (Figure 1.6) 235-237. The β-catenin binding site on E-

cadherin has been mapped to residues, 832-862, as mutation or deletion of this region

abolished the E-cadherin/β-catenin interaction 238. E-cadherin binds to the armadillo

repeats within the central region of β-catenin 239. This interaction first occurs in the

endoplasmic reticulum to stabilize E-cadherin and possibly promote its targeting to the

plasma membrane 240, 241. At the membrane, α-catenin becomes part of the E-cadherin/β-

43

Figure 1.6 Structural components of the E-cadherin-based adherens junction. Cell to cell adhesion is mediated in part by the adherens junction. In epithelial cells E-cadherin is the primary cadherin molecule that facilitates this adhesion. It is composed of an extracellular domain that interacts in a homotypic fashion with cadherins on adjacent cells, a transmembrane region and a cytoplasmic domain that interacts with multiple proteins. p120-catenin binds to a juxtamebrane region of E-cadherin. It serves to recruit different tyrosine kinases, including Fer, to the complex and stabilizes the E-cadherin

44

molecule at the membrane surface. β-catenin binds to the C-terminus of E-cadherin and to α-catenin. The E-cadherin/β-catenin/α-catenin interaction was once thought to be important for the complex to associate with the actin cytoskeleton. Recently this belief has been challenged and a new model has been proposed that suggests that an intact complex is still important for normal cell-cell adhesion but it also has a role in the generation of α-catenin dimers that are proposed to interact with and stabilize the cytoskeleton. In conjunction with tyrosine kinases, tyrosine phosphatases are also present in the adherens junction to regulate the phosphorylation of several components.

45

catenin complex in a 1:1:1 ratio by binding to the N-terminal domain of β-catenin 242.

Because α-catenin can also bind directly to actin filaments 243 it has been popularly

assumed that it provided the link from the adherens junction to the actin cytoskeleton.

Based on this idea it was proposed that α-catenin, β-catenin and E-cadherin formed the

minimal “core complex” required for proper adhesion. However, recent studies have

shown that this may not be true 244, 245. These researchers addressed the question of

whether α-catenin could form a quaternary complex with E-cadherin, β-catenin and actin.

Surprisingly, α-catenin could bind β-catenin or actin but not both simultaneously 244. This

was explained by the ability of α-catenin to exist in two different forms where

monomeric α-catenin preferentially binds the E-cadherin/β-catenin complex while

homodimeric α-catenin binds and bundles actin filaments 244. The α-catenin domains that

are involved in dimerization and binding to β-catenin were shown to overlap and provide

a reason for the mutually exclusive binding 246. Additionally, α-catenin dimers

functionally inhibit the Arp2/3 complex 244 that nucleates branched actin filaments at the

leading edge of cells 247. Normally in the absence of cell-cell contacts, the concentration

of α-catenin in the cytoplasm is insufficient for dimerization to occur, allowing Arp2/3-

mediated actin polymerization to occur 248. However, when cell-cell contacts are formed

the local concentration of α-catenin increases. Due to the weak interaction of α-catenin

and β-catenin there is enough free α-catenin to dimerize, bind actin and locally inhibit

Arp2/3 activity 248. It has been proposed that altering the actin dynamics at cell-cell

contact sites may function to stabilize the cadherin-mediated cell-cell adhesion and

possibly inhibit cell migration 248. A caveat to this model is that an earlier report has

46

described the simultaneous binding of α-catenin to β-catenin and actin 249. Also, the

presence of E-cadherin in the Triton-X insoluble fraction which became soluble after

deletion of the β-catenin binding domain is in contradiction to this model 250. It is likely

that with the ever increasing number of proteins identified within the AJ, and their ability

to interact with multiple partners, there is some form of indirect connection between the

cadherin/catenin complex and the actin cytoskeleton.

p120-catenin also plays important roles in regulating cadherin-based adhesion.

p120-catenin was originally identified as a Src kinase substrate whose phosphorylation

correlated with transformation 251. It was subsequently shown to bind directly to the

cytoplasmic juxtamembrane region of E-cadherin 235, 252. p120-catenin is thought to be

involved in several aspects of AJ function. Prior to AJ formation it promotes the transport

of cadherins to the cell surface 253. Within the AJ, the binding of p120-catenin to

cadherin increases its stability and promotes its retention at the membrane 254, 255. In

SW48 cells, a colon cancer cell line where p120-catenin expression is severely

compromised, E-cadherin localized normally to cell-cell contacts but was not highly

expressed 254. Restoration of p120-catenin expression caused a substantial increase in E-

cadherin levels that was related to increased stability rather than increased expression 254.

In a cell system that expressed normal levels of p120-catenin, the targeting of p120-

catenin by RNAi resulted in a rapid internalization and degradation of E-cadherin and

loss of cell-cell adhesion 255. This was accompanied by the rapid degradation of α- and β-

catenin 255, 256. Thus, the presence of p120-catenin and its interaction with E-cadherin is

47

critical for E-cadherin retention at the cell membrane, along with the establishment and

maintenance of proper cell-cell adhesion.

1.5.2 Targeted disruption of adherens junction components

The role of E-cadherin, α-, β- and p120-catenin in regulating cell-cell adhesion

has important implications for the maintenance of tissue integrity during normal and

disease states. Mutations or alterations in the function of each of these proteins are known

to disrupt normal cell adhesion and contribute to certain types of cancer 257. Not

surprisingly, mice lacking any of the four components are embryonic lethal due to

compromised cellular adhesion 258-261. However, conditional knockout mouse models

have provided the means to study the role(s) of these proteins in adult cells and tissues.

The MMTV/Cre recombinase-mediated inactivation of the E-cadherin gene in the mouse

mammary gland has allowed the function of E-cadherin to be studied during pregnancy

and lactation 262. These mice developed normally up to about mid-pregnancy when the

MMTV promoter becomes active. Starting at day 16 of pregnancy, mutant alveoli

appeared condensed and less developed than normal. By parturition the alveoli contained

small lumina and smaller, misshapen, apoptotic epithelial cells which were more

characteristic of early involution in a normal mouse 262. Obviously, pups from these

mothers died shortly after birth from a lack of nurishment. These results show that E-

cadherin has a role in the terminal differentiation program of epithelial cells during

pregnancy and is essential for survival and function of epithelial cells during lactation.

48

Surprisingly, among the catenins only α-catenin has been deleted in the mammary

gland 263. However, β- and p120-catenin have been deleted in other organs. In the mouse

endothelium the loss of β-catenin caused irregularly shaped vessels that were often

hemorrhagic. Cultured endothelial cells displayed decreased cell-cell adhesion strength

and obvious cytoskeletal changes 264. Other studies with knockout mice have shown that

β-catenin has a role in 1) liver growth and a lesser role in hepatocyte proliferation during

liver regeneration 265; 2) skin stem cell differentiation 266; and 3) brain and craniofacial

development 267. These aspects of β-catenin function may be related to cell-cell adhesion

but they also could be the result of its AJ-independent role in the Wnt signaling

pathway(s) 268 and its ability to act as a transcription factor 269.

Alpha-catenin has been targeted in several organs including the mouse mammary

epithelium, which resulted in blocked epithelial expansion during pregnancy. Targeted

cells lacked proper polarity and differentiation markers and displayed increased apoptosis

at parturition 263. Deletion of α-catenin in cardiomyocytes resulted in 1) reduced

expression of cadherin and vinculin; 2) severe disorganization of intercalated disc

structure; and 3) increased cardiac wall rupture 270. In the skin epidermis, deletion of α-

catenin disrupted adhesion and caused skin cells to become hyperproliferative and highly

invasive 271, 272. Thus, depending on the tissue and type of cell, the deletion of α-catenin

may have very divergent outcomes that may not necessarily be directly related to its role

within the AJ 272.

Deletion of p120-catenin in neonatal epidermis caused a reduction in the AJ

components but no obvious disruption of intercellular adhesion or barrier function.

49

However, with age the mice displayed epidermal hyperplasia and chronic inflammation

273. Ductal branching in the mammary and salivary glands shares some morphological

characteristics and may use a conserved developmental program 274. Therefore, deletion

of p120-catenin in the salivary gland could provide clues as to its function in other

secretory tissue like the mammary gland. When p120-catenin was deleted in the

embryonic salivary gland the development of secretory acini was completely blocked,

resulting in a gland composed entirely of ducts 275. In accordance with its role in cell

adhesion, the loss of p120-catenin resulted in greatly reduced E-cadherin levels and

severe defects in epithelial adhesion, polarity and morphology. Finally, cells within the

ducts were hyperproliferative, leading to tumor-like protrusions that eventually resulted

in ductal blockage 275.

1.5.3 Cadherins, catenins and cancer

Under normal physiological conditions the establishment of E-cadherin-mediated

intercellular adhesive contacts serves to inhibit cell proliferation by a process known as

contact inhibition. It is now commonly accepted that alterations in the adhesion

properties of a cell play a pivotal role in the development and progression of tumors 276.

Being the primary epithelial cell adhesion protein, E-cadherin has been widely studied in

cancer progression. The loss of E-cadherin often occurs during the development of cancer

of various epithelial cell origins. With respect to human breast cancer, E-cadherin

expression is reduced or lost in over 50 % of lobular carcinomas 277. However, other

studies have reported a loss of E-cadherin in 95-100 % of the samples tested 278, 279.

50

Causes of lost expression are varied and include promoter methylation and loss of

heterozygosity as prominent reasons 280-282. Other mechanisms of inhibition include

mutation, transcriptional repression and increased internalization and degradation, but

these are thought to occur less frequently 283-285. There is evidence that the loss of E-

cadherin occurs early in tumorigenesis but other studies suggest that it is not a requisite

for tumor initiation 286, 287. Regardless of the precise timing of its loss E-cadherin is

definitely associated with disease progression and strongly correlates with metastasis 288,

289.

Each of the catenins has been shown to be deregulated in one or more types of

cancer but understanding their role(s) is complicated by the fact that they have important

functions distinct from those associated with intercellular adhesion 273. β-catenin

mutations are commonly found in some tumors, including those of the colon, liver and

ovary 290. Typically, mutations involve a gain-of-function or deletion that serve to

stabilize the protein 290. However, a more recent study has shown that β-catenin

expression is lost in some lobular hyperplasia and carcinomas of the breast 288. Obviously

the presence of β-catenin within the AJ functions to suppress cell growth and its loss

would provide a mechanism to escape the AJ-mediated growth limitations. But β-catenin

exists in at least two distinct pools in cells, one associated with cadherins and another

involved in Wnt signaling and gene transcription 291. Normally in the absence of Wnt

signaling, β-catenin interacts with a protein complex that promotes its phosphorylation,

ubiquitination and degradation 292. When the Wnt pathway is activated the protein

complex that targets β-catenin for degradation is inhibited, allowing the accumulation of

51

β-catenin in the cytoplasm 293. β-catenin then is able to interact with members of the T-

cell factor/lymphoid enhancer factor family of transcription factors and induce the

transcription of certain genes involved in cell growth and invasion 294. Thus, β-catenin

may provide cancerous cells with a growth advantage and explain why it is not frequently

lost during tumor progression. Many types of cancer commonly display a loss or

mutation of the proteins involved in β-catenin degradation 295. Promoter silencing and

mutations in the regulatory proteins that control β-catenin levels are well documented 296,

297.

The role of p120-catenin in cadherin stabilization suggests that it would be lost in

cancerous cells. However, studies have shown this does not occur. Rather it switches

from a membrane to a cytoplasmic and nuclear localization 252. Due to its role in

regulating cadherin stability it has been suggested to precede E-cadherin alteration in

some cancers 252. In support of this, various types of tumors showing altered p120-catenin

staining often display mislocalized or lost E-cadherin staining as well 286, 298, 299. The role

of p120-catenin in the cytoplasm is not known but nuclear p120-catenin interacts with

and inhibits the activity of Kaiso, a sequence-specific and methylation-dependent

transcriptional repressor 300, 301. Kaiso is known to bind the matrilysin promoter and

block its expression 300. Matrilysin, a matrix metalloproteinase, has been implicated in

cancer progression 302. Therefore, a role for p120-catenin may be to inhibit Kaiso from

repressing certain cancer-promoting genes.

Although α-catenin was once considered a simple structural protein linking the

cadherin/catenin complex and the actin cytoskeleton recent studies indicate it is likely to

52

have additional functions which when altered could contribute to cancer progression 303.

Due to its role in cell-cell adhesion it has been assumed that α-catenin would be

frequently mutated or deleted in cancer cells. In support of this idea, α-catenin mutations

have been identified in various cancer cell lines 249, 304, 305. Although human primary

tumors appear to lack similar mutations, α-catenin is often down-regulated or deficient

303. In these cases it is likely that promoter methylation and histone deacetylation are

involved in the reduced α-catenin expression 306. Two lines of evidence suggest that α-

catenin has functions other than in cell-cell adhesion. First, the loss of both α-catenin and

E-cadherin correlates with a prognosis that is worse than the loss of either one alone 307-

309. Second, when compared to E-cadherin, the loss of α-catenin is a stronger indicator of

tumor growth, invasion and metastasis 310-312. Although a definitive role for α-catenin in

cancer has not been identified, it appears to be involved in multiple signaling pathways

that are known to contribute to the development of cancer. Sustained activation of the

Ras/MAPK pathway was shown to contribute to the increased migration and proliferation

observed in the α-catenin-deleted embryonic epidermis 271. Higher proliferation and

reduced apoptosis observed in the central nervous system that lacked α-catenin was

attributed to increased hedgehog and NFκB signaling, respectively 313. Cell culture

studies have shown that over-expression of α-catenin leads to decreased β-catenin

transcriptional activity, whereas the knockdown of α-catenin enhances β-catenin

signaling 314, 315. Finally, α-catenin dimers can inhibit the activity of the Arp2/3 complex

244. If α-catenin does indeed regulate each of these signaling molecules/pathways then it

is possible that its absence would contribute to multiple aspects of cancer progression.

53

1.5.4 Regulation of the adherens junction by tyrosine phosphorylation

Although AJs provide cells with stability and structure, they possess the capacity

to be highly dynamic and rapidly dissociate in response to various stimuli 316. The

phosphorylation of AJ components on tyrosine residues is one of the mechanisms that

regulates the interaction of different components, and ultimately the stability of the

complex 317. Depending on the protein and the site of modification, tyrosine

phosphorylation can have either a positive or negative effect on stability of the protein-

protein interactions and the function of the modified protein. Both receptor and

cytoplasmic tyrosine kinases have been found associated with AJ components. Members

of the epidermal growth factor receptor and Met receptor contribute to the destabilization

of the AJ 318, 319. Cytoplasmic kinases, including members of the Src and Fps/Fer family,

can contribute to both the stabilization and breakdown of the AJ 320, 321.

Not much is known about the extent or function of tyrosine phosphorylation (pY)

of E-cadherin. There are seven conserved tyrosine residues in the cytoplasmic domains of

classical cadherins that are potential phosphorylation sites (Greer, unpublished). A few

reports have described the pY of E-cadherin but none have identified specific sites 322, 323.

Treatment of colon carcinoma cells with acetaldehyde induced E-cadherin pY and

increased paracellular permeability 322. Similarly, HGF stimulation of a breast carcinoma

cell line resulted in a Met receptor-mediated increase in pY of E-cadherin 323. These

results imply that E-cadherin pY may promote disruption of cadherin-based contacts.

54

Conversely, the phosphorylation of three serine residues in the carboxy-terminus of E-

cadherin is critical to its ability to stably interact with β-catenin 238.

Each of the catenins is known to be phosphorylated on one or more tyrosine

residues. The phosphorylation of α-catenin has not been well studied but is known to

occur 324. Although the kinase(s) involved have not been identified, in NIH-3T3 cells

expressing a mutant form of the phosphatase, SHP2, the pY of α-catenin was shown to

enhance its translocation to the plasma membrane and interaction with β-catenin, leading

to the stabilization of intercellular adhesion 324. It is not known if this type of mechanism

is required for normal cell-cell adhesion or if it is even common to other types of cells.

The phosphorylation of β-catenin and its role(s) in regulating β-catenin interaction

with other AJ components is better understood. Three conserved tyrosines, 142, 489 and

654, within β-catenin can be phosphorylated 325. Phosphorylation of β-catenin is

generally associated with a downregulation of cadherin-mediated adhesion. Tyrosine 142

is in the region that binds α-catenin 320. Phosphorylation of this residue causes the

dissociation of α-catenin from β-catenin and reduced cell-cell adhesion 326. Both Fyn and

Fer kinases, and possibly c-Met, can mediate tyrosine 142 phosphorylation 320, 325. While

the exact biological processes are not well understood, Fyn- and Fer-mediated β-catenin

pY142 is known to occur in response to cell shrinkage 327and upon K-ras activation 320.

This type of mechanism may have a role in cytoskeletal reorganization, as well as

reducing cell adhesion during tumor development.

While the phosphorylation of both Y489 and Y654 causes the release of β-catenin

from cadherin 328, 329, Y654 phosphorylation seems to be more prevalent. Y654 can be

55

phosphorylated by the Src, ErbB2 and c-Met kinases 328, 330, 331. The fact that each kinase

is commonly deregulated in cancer cells suggests that the phosphorylation of β-catenin at

Y654 is likely a key step in this process. The regulation of Y654 phosphorylation also

involves the tyrosine phosphatase, PTP1B, which is bound directly to cadherin 321, 332. An

interesting aspect of PTP1B function is that its interaction with cadherin depends on the

phosphorylation on Y152 of PTP1B which is mediated by Fer 321, 333. Fer itself is part of

the cadherin complex through its association with p120-catenin 334. The mutation of Fer

in fibroblasts expressing N-cadherin results in the progressive loss of PTP1B, β-catenin

and p120-catenin from the AJ, with a concomitant increase in β-catenin phosphorylation

and decrease in cell-cell adhesion 321. Thus, the Fer kinase has a somewhat paradoxical

relationship with β-catenin in that it is involved in both its phosphorylation and

dephosphorylation.

In addition to its cadherin-stabilization function, p120-catenin serves as a

scaffolding protein to recruit different tyrosine kinases into the AJ. The Src-family

kinases, Yes and Fyn, as well as the Fer kinase can interact directly with p120-catenin 320.

Much of the tyrosine phosphorylation of the different AJ components is due to the ability

of p120-catenin to interact with and recruit different kinases. Under certain conditions

p120-catenin itself can be highly phosphorylated. In fact, it contains approximately

twenty tyrosine residues that are known sites of phosphorylation. Eight of these residues

were shown to be v-Src substrates 335 and multiple p120-catenin tyrosines were

phosphorylated in v-Src transformed cells 336. Activation of the EGFR also induces p120-

catenin phosphorylation but it is not known if intermediate kinases are involved 337. The

56

over-expression of Fer caused an increase in overall p120-catenin pY content 334.

Similarly, cell shrinkage induced p120-catenin phosphorylation that was dependent on

the Fyn kinase 327. Despite these observations, the functional consequences of

phosphorylation are generally not well understood. In some instances p120-catenin

phosphorylation correlated with an increased affinity for cadherin 320, 338. Conversely, the

association between p120-catenin and E-cadherin was not altered in v-Src transformed or

receptor tyrosine kinase stimulated cells 235, 337, 339. In some cases specific phosphorylated

tyrosines are thought to serve as docking sites for certain interacting proteins. This is true

for SHP-1, a cytoplasmic phosphatase that binds to pY on p120-catenin through its SH2

domain 340. Also, the interaction of the RhoA GTPase with p120-catenin is regulated by

specific pY residues 341. Depending on the site of phosphorylation the affinity of p120-

catenin toward RhoA can be positively or negatively affected. Fyn-mediated

phosphorylation of Y112 inhibits the interaction with RhoA whereas the phosphorylation

of Y217 and Y228 by Src promotes binding to RhoA 341. p120-catenin is known to inhibit

RhoA activity, while increasing that of Rac1 and Cdc42, and may shuttle between

cadherins and the cytoplasm where it acts on these GTPases 342, 343. In this manner it is

likely that p120-catenin is involved in the dynamic remodeling of the actin cytoskeleton

that occurs during migration and invasion 344. Together, these types of results probably

reflect the complex nature of p120-catenin function and the effect that individual or

combinations of tyrosine phosphorylation have on the ability of p120-catenin to interact

with binding partners and regulate their function.

57

1.6 Hypothesis and research objectives

In a study by the Leder lab the expression of over forty different tyrosine kinases

was analyzed during mouse mammary development and revealed that fps transcript levels

were upregulated late in pregnancy and increased further during lactation 107. A

subsequent report identified Fps as a possible tumor suppressor in the mouse mammary

gland 55. These data have led to the development of my hypothesis that the Fps tyrosine

kinase is a component of one or more signaling pathways which regulate the normal

function of the mammary gland. Comparison of the physiological and biochemical

differences between the wild-type and fps-null mice should elucidate the function of the

Fps kinase in the mammary gland.

The first objective of my thesis was to characterize the function of the Fps

tyrosine kinase in the mammary gland. Initially the expression of the Fps protein was

examined during different stages of mammary gland development. As a measure of its

kinase activity, the phosphotyrosine content of Fps was also assessed during the same

developmental stages to determine if levels of activity coincided with protein expression.

Confocal microscopy was used to identify the expression pattern of Fps. Differences in

pup weights from wild-type and fps-null matings led us to examine the mammary glands

by whole mount and histological analyses. The protein and fat content of the milk from

both genotypes was assessed. Potential substrates and interacting partners were studied

by immunoprecipitation and/or immunoblotting. The results obtained here suggest that

Fps is most highly expressed and activated during lactation, and is important for proper

mammary gland homeostasis and function. The presence of Fps in the E-cadherin-based

58

adherens junction protein complex suggests that Fps may regulate cell-cell adhesion and

its absence could compromise the integrity of the gland.

The second objective of my thesis was to establish a cell culture based system to

study the function of Fps in mammary epithelial cells. First, culture conditions were

established to generate enriched populations of mammary epithelial cells. The primary

cells from a fps-null gland were then infected with a lentivirus encoding a p53 mouse-

specific short hairpin RNA. Continued growth of infected cells either in or on a collagen

gel matrix resulted in the outgrowth of a cell population that displayed characteristics of

an immortalized epithelial cell line. Subsequent infection of cells with a virus expressing

wild-type Fps has provided the reagents to study the function of Fps in breast epithelial

cells and determine the phenotype(s) associated with its absence.

59

Chapter 2

The Fps tyrosine kinase is up-regulated during lactation in the mouse mammary

gland and is a component of the E-cadherin adherens junction.

2.1 Abstract

Fps is a 92 kDa cytoplasmic protein-tyrosine kinase. It is thought to regulate

diverse cellular functions such as the immune response and cell differentiation, and has

been reported to act as a tumor suppressor in the mouse mammary gland. However, its

role in normal mammary gland development has not been addressed. Here we report that

the protein levels of Fps increased during pregnancy and reached a maximum during

lactation. There was also a direct correlation between the level of expression and the

extent of Fps tyrosine phosphorylation. Pups reared by fps-null females gained weight

more slowly than pups reared by wild-type females. Fps expression was observed in

epithelial cells and localized throughout the cytoplasm in a punctate pattern. No

differences in the protein or fat content of the milk were detected. Neither whole mount

nor histological analyses of mammary glands from different developmental stages

showed any morphological defects in fps-null females. Fps was identified in a complex

with E-cadherin, β-catenin and p120-catenin predominantly during lactation.

Immunofluorescence analysis indicated that E-cadherin and β-catenin localization was

less concentrated at sites of cell-cell contacts in the fps-null glands. Analysis of the

interactions between different adherens junction components suggests that less E-

60

cadherin and β-catenin was associated with p120-catenin in the fps-null glands. No

differences in the phosphotyrosine status of the different adherens junction components

were detected.

2.2 Introduction

The primary function of the mammary gland is the synthesis and secretion of milk

to nurture the young. In the adult female, the mammary gland undergoes a cycle of cell

proliferation, differentiation, milk production and regression with each round of

reproduction. This process depends on a highly ordered series of events involving both

spatial and temporal interactions among several different mammary cell types that are

themselves coordinated by a combination of growth factors and steroid hormones 345.

Protein kinases regulate mammary gland development, differentiation and carcinogenesis

346 and the expression and function of many protein kinases are known to be dynamically

regulated during normal mammary gland development 107. Consequently, any alteration

in the proper expression or activity of a protein kinase may result in abnormal

development or function of the mammary gland, and may also contribute to

tumorigenesis.

The fps/fes proto-oncogene encodes a 92 kDa cytoplasmic tyrosine kinase

(hereafter referred to as Fps). Fps is composed of a putative amino terminal F-BAR

domain (comprised of a Fps/CIP4 homology (FCH) motif and a coiled-coil motif),

followed by a second predicted coiled-coil (CC) motif, a Src homology-2 (SH2) domain,

and a carboxy-terminal tyrosine kinase domain 5, 66, 67, 87. Fps shares close structural

61

similarity with a 94 kDa protein called Fer (Fes related) 36 and together make up a distinct

subclass within the non-receptor tyrosine kinase family 65. While Fer appears to be

expressed ubiquitously, Fps expression is somewhat more restricted. To date, many of the

cell types within the hematopoietic system (reviewed in 5, 87) and some types of

endothelial 51, neuronal and epithelial cells have been shown to express Fps 52, 347, 348.

Fps/Fes was first isolated as an oncoprotein encoded by tumor-causing avian

(Fps) and feline (Fes) retroviruses 2. These oncogenic proteins consisted of amino-

terminal retroviral Gag sequences fused to some or all of the Fps/Fes sequence, and this

contributed to unregulated tyrosine kinase activity. In contrast, the activity of normal

cellular Fps is tightly regulated 87 and induces little or no transformation when ectopically

expressed in rodent fibroblasts but can transform certain cell types if expressed at

sufficient levels 349or expressed as an amino-terminal myristoylated variant 51, 350.

Regulation of Fps kinase activity is controlled, at least in part, by N-terminal sequences

86. This is supported by the fact that all viral oncogenic forms are fusions with the N-

terminus of Fps 5 and a point mutation of the first coiled-coil motif dramatically increased

kinase and transforming activities of Fps 86, 351. No oncogenic alleles of Fer are known;

however, Fer possesses a much higher intrinsic kinase activity than Fps and can induce

transformation upon ectopic expression 352.

The role of Fps has been studied predominantly in cytokine signaling. Tyrosine

phosphorylation of Fps has been observed in different cell types of the hematopoietic

system after treatment with granulocyte/monocyte colony-stimulating factor,

erythropoietin, stem cell factor, and several interleukins (IL-3, IL-4, IL-5, IL-6),

62

suggesting that Fps is a component of a large number of signal transduction pathways

(reviewed in 5, 87). Several in vitro studies have pointed to Fps as a positive regulator of

cell differentiation, specifically for macrophage and granulocytic cells 94, 353-357. In

contrast to these observations from cell culture systems are the results generated from

transgenic mouse studies where neither a targeted fps kinase-inactivating mutation nor a

fps-null mutation resulted in any dramatic in vivo steady-state hematopoietic phenotypes

26, 101. Fps may also promote differentiation of neuronal cells, as expression of Fps in

PC12 cells accentuated the rate and length of process extension 348. Through fluorescence

microscopy analysis, Fps co-localized with the trans golgi network proteins, TGN38 and

γ-adaptin, and with Rab proteins involved in endocytosis (Rab5B and Rab7) or

exocytosis (Rab1A and Rab3A) 52, 63, suggesting a possible role for Fps in vesicle

trafficking. In support of this is the observation that Fps regulates internalization and

down-regulation of the Toll-like receptor 4 102.

E-cadherin is a member of the type I classical cadherin family of calcium-

dependent intercellular adhesion molecules 358. It is the prominent cadherin expressed in

most types of epithelial cells and mediates cell-cell adhesion through homophilic

interactions between extracellular domains on adjacent cells 233. Clustering of E-cadherin

molecules into localized zones, or adherens junctions (AJ), is essential for strong cell-cell

adhesion and normal epithelial characteristics, including establishment of epithelial cell

shape, maintenance of a differentiated epithelial phenotype and cell polarization 359

(reviewed in 360). The cytoplasmic domain of E-cadherin associates either directly or

indirectly with several intracellular proteins, including members of the catenin family. In

63

particular, p120-catenin interacts with E-cadherin and regulates its stability and retention

at the cell surface, with its absence or down-regulation resulting in cadherin

internalization and degradation 254, 255. p120-catenin also serves to recruit several

different tyrosine kinases to the AJ 320, 321. β-catenin binds E-cadherin and α-catenin

which, in turn, establishes a link with the actin cytoskeleton 243. These proteins are

thought to represent the core complex of the AJ, required for strong cell-cell adhesion 361.

However, recent evidence indicates that α-catenin can bind to E-cadherin-bound β-

catenin or actin, but not both simultaneously 244, 245; suggesting that the cadherin-

cytoskeleton link is more dynamic and complex than previously believed.

Disruption of any of the AJ core proteins can lead to decreased cell-cell adhesion

and altered function 264, 271, 275. Loss or mutation of the different components has been

reported in multiple cancers types 362-364. Deregulation of E-cadherin is thought to disrupt

its ability to function as a suppressor of tumor growth and invasion 286, 288, 289.

Conversely, in the differentiating mouse mammary gland, conditional inactivation of the

E-cadherin gene severely affected the terminal differentiation program of the mammary

alveolar epithelial cells and resulted in massive cell death at parturition 262. Deletion of β-

catenin in different cell types indicates that it is involved in regulating cell adhesion and

the cytoskeleton 264, cell growth 265, differentiation 266 and tissue development 267. Over-

expression of a stabilized form of β-catenin in the mouse mammary epithelium led to

premature cell differentiation during pregnancy and increased cell proliferation that

persisted into lactation 365. Mutant forms of β-catenin that do not bind to cadherin or have

lost the domain required for targeted degradation have been identified in some types of

64

cancer 290. Targeted deletion of p120-catenin in the salivary gland caused defects in

epithelial adhesion, polarity and morphology that blocked acini development and led to

hyperplasia 275. The expression of p120-catenin is typically maintained in cancer cells but

shifts to a cytoplasmic or nuclear localization 252 where it may provide a growth

advantage.

It is well documented that phosphorylation is an important mechanism in

regulating the interaction of AJ proteins and the stability of the core complex 321, 322, 338.

Tyrosine phosphorylation of AJ components can have different effects depending on the

protein and the site of modification. Under some conditions phosphorylation of p120-

catenin is thought to increase its affinity for cadherins 320, 366 but is not essential for this

interaction 235, 337. Similarly, specific phosphorylation sites of p120-catenin may act as

docking sites for other interacting proteins 340, 341. The phosphorylation of p120-catenin

has also been linked with cell transformation 366.

In general, the tyrosine phosphorylation of β-catenin is associated with reduced

cell-cell adhesion 325. The phosphorylation of Y142 and Y654 disrupts its interaction with

α-catenin and cadherin, respectively, and promotes disassembly of the adherens junction

complex 320, 328, 334. Several protein tyrosine kinases and phosphatases have been

identified in adherens junctions. The Src-family kinases, Src and Fyn, are known to

phosphorylate p120-catenin and β-catenin under a variety of different stimuli (reviewed

in 317, 325. The Fer kinase can interact with and phosphorylate p120-catenin and β-catenin

as well 320, 321, 334, 367. Fer is known to phosphorylate β-catenin on Y142 320. Fer can also

phosphorylate the tyrosine phosphatase, PTP1B 321, on Y152 which promotes its

65

interaction with N-cadherin 333. In turn, PTP1B serves to maintain Y654 of β-catenin in

an unphosphorylated state and stabilize the AJ 321. Together, the activities of Fer enable it

to regulate both phosphorylation and dephosphorylation of AJ components. Over-

expression of Fer in embryonic fibroblasts that expressed E-cadherin resulted in the

rounding up and detachment of cells that was the result of the Fer-induced

phosphorylation and dissociation of β-catenin from the E-cadherin complex 334. The loss

of Fer activity in mouse embryonic fibroblasts resulted in the dissociation of PTP1B and

β-catenin from N-cadherin as well as the loss of N-cadherin and β-catenin from cell-cell

contacts 321.

A survey to identify protein kinases expressed in the mammary gland and to

examine their expression patterns demonstrated that Fps mRNA levels are differentially

expressed during mammary development 107. This was the first study to suggest that Fps

may have a function in the mammary gland. More recently, our lab has shown that in a

mouse mammary tumor virus (MMTV)/polyoma middle T mouse model for breast

cancer the loss or inactivation of Fps resulted in the development of tumors with an

earlier onset than in a wild-type background 55, suggesting that Fps may possess tumor

suppressor abilities 53-55.

In this study the role of the Fps tyrosine kinase in the development and function

of the mouse mammary gland has been investigated. We have demonstrated that the

levels of the Fps protein rise during pregnancy and reach maximal expression during

lactation. This was accompanied by a substantial increase in the tyrosine phosphorylation

of Fps during lactation. Pups reared by fps-null mothers gained weight more slowly than

66

wild-type counterparts. The expression of Fps was identified in epithelial cells. Whole

mount and histological analyses did not identify any obvious differences in the overall

ductal morphology. A search for interacting partners and potential substrates revealed

that both Fps and Fer are components of the E-cadherin-based AJ complex. Despite this

interaction, no differences in the phosphotyrosine status were observed in the core

adherens junction components in fps-null tissue. However, less E-cadherin and β-catenin

appeared to be associated with p120-catenin in fps-null tissue. Fluorescence microscopy

revealed that the staining of E-cadherin and β-catenin in fps-null tissue was more

disorganized at sites of cell-cell contacts. Together, these results point to the Fps kinase

as a potential regulator of epithelial cell AJs and lactation.

2.3 Materials and Methods

2.3.1 Mice

The fps-null and c-fes mouse lines have been described previously 26, 28. All

mouse lines were maintained in an in-bred SVJ/129-CD1 hybrid background and housed

in the Animal Care Facility at Queen’s University, Kingston, Ontario, Canada. For the

isolation of stage-specific mammary glands, females were mated at 8-9 weeks of age and

the day of vaginal plug formation was considered as day 0. The third, fourth and fifth

mammary glands were isolated for biochemical analysis, with the lymph node of the

fourth gland removed prior to analysis. The fourth gland was used for all other

experiments.

67

2.3.2 Pup weight analysis

Breeding pairs were established and monitored for successful matings. Beginning

one day after birth the weight of each pup was measured at approximately the same time

daily. All statistical values were calculated using the Student’s t-test. All error bars

represent the standard error of the mean.

2.3.3 Antibodies

The antibodies to Fps/Fer (FpsQE) and Fer (FerLA) have been described

previously 52. The following antibodies (supplier) were used; E-cadherin, α-catenin, β-

catenin, p120-catenin, Dynamin II, p130Cas and Fak (BD Transduction Labs), PY99,

Erk1/2 and actin (Santa Cruz), cortactin, Stat5A and pStat5A/B (Upstate Biotechnology),

smooth muscle actin (SMA) and α-tubulin (Sigma-Aldrich), Mucin-1 (Neomarkers), NSF

(Stressgen Biotechnologies), Akt (Cell Signaling Technology). Antibodies to Pacsin-2

and p190RhoGAP were provided by Dr. Markus Plomann and Dr. Sarah Parsons,

respectively. The p85 antibody was a gift from Dr. Jane McGlade. All α-rabbit and α-

mouse secondary antibodies used in the immunofluorescence experiments were obtained

from Molecular Probes. Alexa-633 and Alexa-533 conjugated IgG F(ab)2 fragments were

used for red and green fluorescence, respectively. The FITC-labeled α-armenian hamster

secondary antibody was purchased from eBiosciences.

68

2.3.4 Immunoblotting, immunoprecipitation and kinase assay analysis

For immunoblot analysis, protein lysates from mammary glands were prepared in

PBS/NP-40 buffer (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, 1 %

Nonidet-40, [pH 7.3] with 10 μg/ml aprotinin, 10 μg/ml leupeptin, 100 μM sodium

orthovanadate, 100 μM phenylmethylsulfonyl fluoride) using an Ultra-Turrax T25

homogenizer (Terochem Scientific). For analysis of nuclear proteins the tissue was

homogenized in radioimmunoprecipitation assay buffer (50 mM Tris-HCl, [pH 7.4], 150

mM NaCl, 1 mM EDTA, 1 % Triton X-100, 1 % sodium deoxycholate, 0.1 % SDS, with

inhibitors), followed by two 20 second bursts of sonication. Cell lysates were centrifuged

at 14000 x g for 15 minutes at 4oC and quantified using the Bio-Rad protein assay (Bio-

Rad). Lysates were diluted to 1 mg/ml with 2x sodium dodecyl sulfate sample buffer

(130 mM Tris-HCl [pH 6.8], 20 % [v/v] glycerol, 2 % [w/v] sodium dodecyl sulfate, 2 %

[v/v] β-mercaptoethanol, 0.08 % [w/v] bromophenol blue) and heated for 5 minutes at

95oC. Aliquots of 10 μg were separated by sodium dodecyl sulfate polyacrylamide gel

electrophoresis (SDS-PAGE) and transferred to Immobilon P membranes (Millipore)

using a semidry transfer apparatus (Bio-Rad). Membranes were blocked with 5 %

skimmed milk in TBST buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1 % Tween-

20) for rabbit antibodies and 5 % BSA in TBST for mouse antibodies, and then incubated

overnight at 4oC with the specified antibody. After washing in TBST, membranes were

incubated with a 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit

immunoglobulin G (IgG) (Vector Laboratories), or goat anti-mouse IgG (GE Healthcare)

69

in TBST for 1 hr at RT. After washing with TBST, reactive proteins were detected by

using enhanced chemiluminescence reagents (NEN Life Science Products).

For immunoprecipitation (IP), each cell lysate containing 1 mg of protein was

diluted to 1 ml with PBS/NP-40 buffer and incubated O/N at 4oC with the 2-3 μg of

antibody and 30 μl of 50 % (v/v) gammabind sepharose beads (Amersham).

Immunocomplexes were washed three times with PBS/NP-40 buffer and heated for 5

minutes at 95oC in 2x SDS buffer. Following separation by SDS-PAGE, proteins were

transferred to membrane and probed with the specified antibody. Kinase assays were

performed on IPs where, after the third wash, the immunocomplexes were washed once

with Kinase Reaction Buffer (20 mM Tris-HCl [pH 7.5], 10 mM MnCl2, and 100 μM

sodium orthovanadate) and then incubated at 30oC for 20 minutes in KRB with 0.5 mM

ATP.

2.3.5 Whole mount and histological analysis

For whole mount analysis, fresh mammary glands were spread on glass

microscope slides and allowed to adhere for 1 hr. The glands were incubated in acetone

O/N to remove the fat and then treated with 100 % and 95 % ethanol for 1 hr each

followed by hematoxylin staining (10 % hematoxylin [w/v] in 95 % ethanol) for 2 hrs.

Samples were rinsed in tap water for up to 1 hr and destained with 25 mM HCl in 50 %

ethanol until the background staining was clear. Following dehydration with ethanol and

clearing with xylene, each sample was overlaid with Permount mounting media and

covered with a glass coverslip.

70

For histological analysis, mammary tissues were fixed O/N at 4oC in 10 %

formalin in PBS. Sample processing and histological staining was performed by Mr. John

Dacosta, Department of Pathology and Molecular Medicine, Queen’s University. In brief,

the tissue was dehydrated through an ethanol series, cleared with xylene and embedded in

paraffin wax. The tissue was then cut into 8 μm sections and stained with hematoxylin

and eosin. Photos were captured on a Nikon Eclipse E600 microscope fitted with a DXM

1200 digital camera.

2.3.6 Immunofluorescence

Paraffin embedded tissue sections were deparaffinized with xylene or Citrisolv

(Fisher) and rehydrated through an ethanol series. Antigen retrieval was performed by

heating the samples in 10 mM sodium citrate pH 6.0 in a microwave for 8 to 10 minutes

and allowed to cool for 30 minutes. Following 3 washes in PBS, the samples were

blocked for 1 hr with 5 % goat serum in PBS/ 0.02 % Triton X-100. All primary

antibodies were diluted 1:50 in blocking buffer and incubated with the tissue O/N at 4oC.

The samples were then washed 3 times in PBS before addition of the appropriate Alexa

fluor-conjugated secondary IgG F(ab)2 fragment (1:200 dilution in blocking buffer) for 1

hr at RT. For the Mucin-1 antibody, a FITC-conjugated α-armenian hamster IgG

antibody was used. Nuclei were stained with 2 mM Hoechst 33258 for 10 minutes in

PBS. Sections were washed 3 times for 10 minutes each with PBS and then mounted with

Mowiol (Calbiochem). Images were captured on a Leica SP2 true scanning confocal

microscope.

71

2.3.7 Milk protein and fat analysis

For milk collection, the mothers were separated from the pups two hrs before

injection with 0.2 IU oxytocin (Sigma). Fifteen minutes after injection the mice were

anesthetized with 250 mg/kg Avertin. The glands were gently massaged to force the milk

toward the nipple. The milk was then squeezed into a 1.5 ml eppendorf tube. To compare

the expression of the major milk proteins between the two genotypes, the protein

concentration was determined and 5 μg were diluted with 2x SDS sample buffer,

separated by SDS-PAGE and stained with Coomassie blue. The total protein

concentration of the milk was determined using a standard Bio-Rad protein assay (Bio-

Rad). For total milk fat analysis, the concentration of triglycerides was measured using

the Wako L-Type TG-H assay according to the manufacturers’ instructions (Wako

Chemicals USA). Error bars indicate the standard error of the mean.

2.4 Results

2.4.1 Expression of the Fps tyrosine kinase increases during pregnancy and reaches

maximal levels during lactation.

An analysis using RT-PCR indicated that mRNA levels of the Fps kinase were

up-regulated during late pregnancy and lactation 107. A more recent study suggested that

Fps regulates tumor formation in the mouse mammary gland by acting as a tumor

suppressor 55. As an initial step towards understanding the function of Fps in the

mammary gland we have examined its protein expression during different stages of

mammary development. Mammary glands were harvested from mice at 3, 6 and 9 weeks

72

of age, day 6, 12 and 18 of pregnancy, day 4 and 12 of lactation, and day 2 and 7 of

involution. Equal amounts of total protein lysates were then assessed by immunoblotting

(Figure 2.1). Using either an anti-Fps/Fer antibody (Figure 2.1, upper panel) or an anti-

Fer antibody (Figure 2.1, middle panel) the expression of Fer remained relatively

constant over all times examined. In contrast, Fps expression was differentially expressed

(Figure 2.1, upper panel). During the first 9 weeks of age and early pregnancy the level

of Fps expression was approximately equal to Fer. However, in late pregnancy, Fps

expression increased relative to Fer, and increased further to reach maximum levels

during lactation before declining to pre-pregnancy levels during involution. The

decreased levels of actin, as a proportion of total protein, which were observed during

late pregnancy and lactation was presumably due to an increase in the relative levels of

milk proteins (Figure 2.1, lower panel).

2.4.2 The Fps tyrosine kinase is highly phosphorylated during lactation.

In its inactive state, Fps is unphosphorylated and becomes phosphorylated on two

or three key tyrosine residues upon activation 94, 349. To address whether the increase in

Fps expression correlated with activity, the phosphotyrosine status of Fps was

determined. Only the stages of pregnancy and lactation where Fps expression increased

were studied. Immunoblotting analysis of Fps/Fer IPs with an anti-phosphotyrosine

antibody showed that Fps became activated during late pregnancy, followed by a

dramatic increase in activation during lactation (Figure 2.2A). These data show that the

increase in expression is associated with a concomitant increase in kinase activity. To

73

Figure 2.1 Expression of the Fps tyrosine kinase during mammary development. For each stage of development mammary glands were harvested from three mice and equal amounts of protein from each sample were pooled. Aliquots of 10 μg were analyzed by immunoblot analysis (IB) using the α-Fps/Fer antibody that detects both Fps and Fer proteins. Parallel experiments were performed with α-Fer and α-actin antibodies. wks (weeks after birth), Preg (pregnancy), Lact (lactation), Inv (involution), D (day).

74

Figure 2.2 The Fps tyrosine kinase is highly activated in the mammary gland during lactation. (A) Wild-type mammary gland lysates from different stages of pregnancy and lactation were subjected to α-Fps/Fer immunoprecipitation (IP) and immunoblotted (IB) with α-pY99 to determine the phosphotyrosine content. (B) The phosphotyrosine content of α-Fps/Fer immunoprecipitations from both wild-type (WT) and fps-null (N) lactating glands was compared to that of Rat-2 cell lines that over-express Fps (F) or constitutively activated myristoylated Fps (MF). SCL (Soluble cell lysate).

75

eliminate the possibility that Fer was activated instead of Fps, the same analysis was

performed on lactating mammary glands from wild-type and fps-null mice. As expected a

distinct phosphotyrosine band representing Fps was observed only in the wild-type

sample (Figure 2.2B). The level of Fps phosphorylation in the lactating mammary gland

was also compared to that in the brain, lung, liver, pancreas, spleen, gut, abdominal fat

and muscle from a wild-type mouse. Only the sample from the mammary gland contained

a phosphotyrosine band that corresponded to Fps (data not shown). To determine whether

the observed level of Fps tyrosine phosphorylation was merely related to increased

expression or to an actual increase in activity, the phosphotyrosine level of Fps in the

mammary gland was compared to the levels found in two Rat-2 cell lines that ectopically

express either wild type cellular Fps (F) or an activated myristoylated Fps variant (MF),

respectively 51. In the cells that over-express wild-type Fps, Fps was not detectably

autophosphorylated in vivo on tyrosine. In contrast, the MF protein was tyrosine

phosphorylated, and this correlated with a transformed morphology of these cells relative

to cells expressing wild type Fps. By comparing the intensity of the phosphotyrosine

bands to the total level of Fps protein in each IP, it is apparent that wild-type Fps in

lactating breast is even more active than the active MF variant in transformed cells

(Figure 2.2B).

76

2.4.3 Offspring raised by fps-null females gained weight more slowly than those raised

by wild-type females.

The observations that Fps expression and kinase activity were maximal during

lactation suggested that Fps may be important to mammary gland function during pup

rearing. To address this possibility, combinations of breeding pairs with wild-type and

fps-null genetic backgrounds were established and the weights of their pups were

measured daily from birth to weaning at day 20. Pups from ♂ +/+ x ♀ +/+ breeding pairs

gained weight significantly faster (p< 1.27 x 10-7 for all time points) than pups from

either ♂ -/- x ♀ -/- or ♂ +/+ x ♀ -/- breeding pairs (Figure 2.3A). There was no

difference in the weight gains of fps-null pups or heterozygous pups reared by fps-null

females in the latter two sets of breeding pairs, which supported the hypothesis that the

pup weight gain defect was not associated with the genotypes of pups, but rather that of

the mothers. This was further supported by measurements of pup weight gains as a

function of their genotypes in litters from heterozygous breeding pairs (Figure 2.3B).

When reared by heterozygous females, pups of all three genotypes (+/+, +/- or -/-) gained

weight at the same rates (Figure 2.3B). Interestingly, pups reared by fps-null females that

over-expressed the human c-fes transgene gained significantly more weight (after day 4

p< 0.03 for all time points) than those reared by fps-null females (Figure 2.3A). Taken

together, these data strongly implicate Fps in the regulation of mammary gland function

during lactation.

77

Figure 2.3 Analysis of pup weights. (A) Wild-type (+/+), fps-null (-/-) or c-fes rescue (c-fes;-/-) mice were set up as breeding pairs, and the weights of the pups were recorded daily from the day after birth (d1) to weaning age (d20). (B) Heterozygous breeding pairs (+/- x +/-) were set up and the weights of the resulting wild type (+/+), heterozygous (+/-) and homozygous mutant (-/-) pups were recorded as in panel A. The genotype of the pups did not influence their rate of weight gain, but pups reared by fps-null mothers gained weight significantly more slowly than those reared by wild-type mothers (p<1.27 x 10-7), and the c-fes rescue transgene partially corrected this weight gain defect (p< 0.03). N > 34 for all time points. Error bars represent the standard error of the mean.

78

2.4.4 The Fps tyrosine kinase is expressed in mammary epithelial cells.

Because the mammary gland is composed of numerous types of cells and Fps is

not ubiquitously expressed, it was important to identify its cellular expression pattern in

the lactating gland. With the use of cell-lineage specific markers, the expression of Fps

was analyzed by immunofluorescence. Mucin-1 is prominently expressed on the apical

surface of differentiated epithelial cells and was used as a mammary epithelial marker 368.

By comparing the staining pattern of the Fps/Fer antibody with that of Mucin-1, Fps

expression was identified in both ductal (not shown) and luminal alveolar epithelial cells

(Figure 2.4A). A similar comparison with the expression pattern of a myoepithelial

marker, smooth muscle actin (SMA) 369, showed that Fps was not expressed in this cell

type (Figure 2.4B). Although other types of cells, including endothelial cells, are known

to express Fps and may contribute to its overall mammary gland expression, based upon

these immunofluorescence observations we believe the epithelial cells are the primary

source of Fps expression.

2.4.5 Milk from fps-null mice has normal protein and fat content.

Although no obvious defects in the overall development of the mammary gland

were observed in fps-null mice, it was possible that the milk from fps-null females was

qualitatively altered. However, when milk proteins were compared by Coomassie blue

staining there was no reduction in the relative intensities of the major milk proteins in the

fps-null samples (Figure 2.5A). Similarly, when the total protein and triglyceride

concentrations were determined no differences were observed (Figure 2.5B and C).

79

Figure 2.4 Cell-specific expression of the Fps tyrosine kinase in the mammary gland during lactation. Paraffin sections of wild-type and fps-null mammary glands at day12 of lactation were stained for Fps/Fer (red) and the epithelial cell marker, Mucin-1 (green) (A), or the myoepithelial cell marker, SMA (green) (B). The nuclei were viewed with Hoechst 33258 (blue).

80

Figure 2.5 Analysis of milk proteins and triglycerides. (A) Equal volumes of milk from wild-type and fps-null mice were separated by SDS-PAGE and stained with Coomassie blue. (B) The concentration of total protein was determined using a standard Bio-Rad assay, and (C) the triglycerides were quantified using the Wako L-Type TG-H assay. For (B) and (C) N = 10. Error bars represent the standard error of the mean.

81

2.4.6 Analysis of the expression and phosphotyrosine content of various signaling

molecules in the mammary gland.

The suspected compromised mammary gland of the fps-null mice suggested that the

activity of one or more signaling molecule was altered. The level of expression and

phosphotyrosine status of Stat5A/B, N-ethylmaleimide-sensitive factor (NSF), Pacsin-2,

Dynamin II, Fak, p85, p190RhoGap, cortactin, p130Cas, suggests that Fps does not

directly regulate the expression or activity of these proteins during lactation (Figure 2.6).

Akt and Erk were also analyzed but a phosphorylation signal was not detected for either

protein.

2.4.7 The morphology of the fps-null mammary gland is normal.

The lower rate of weight gain of pups reared by fps-null females suggested that

mammary development might be compromised in the fps-null females. To explore this

possibility, mammary glands from wild-type and fps-null mice were isolated at different

stages of development and analyzed by whole mount hematoxylin staining. Analysis of

tissues from virgin, day 14 of pregnancy and day 4 of lactation did not show any

differences in the pattern or number of ductal extensions between the wild-type and fps-

null breast tissues (Figure 2.7A). A more in depth examination of the alveolar structures

during lactation was performed on H&E stained sections from wild-type and fps-null

tissue. No significant differences in the size, number or architecture were observed in the

alveoli of fps-null tissue (Figure 2.7B).

82

Figure 2.6 Immunoblotting analysis of potential substrates of the Fps tyrosine kinase. Using 1mg of total protein lysate from wild-type and fps-null lactating mammary glands immunoprecipitations (IP) were performed with the indicated antibodies, followed by immunoblotting (IB) with the indicated antibodies. For direct immunoblotting analysis, 10 μg of soluble protein was probed with the specified antibodies.

83

84

Figure 2.7 Whole mount and histological analysis of wild-type and fps-null mammary glands. (A) Whole glands from pregnant (Day 14) and lactating (Day 12) mice were stained with hematoxylin to visualize the ductal structures. (B) Paraffin sections of glands during lactation (Day 12) were stained with hematoxylin and eosin to compare the alveolar structures in wild-type and fps-null mice. Shown are low, medium and high magnifications in the top, middle and bottom panels. Scale bar (80 μm upper, 40 μm middle, 20 μm lower).

85

2.4.8 The Fps and Fer tyrosine kinases are components of the E-cadherin-based

adherens junction in the mammary gland during lactation.

The Fps-related Fer kinase is known to be a component of both N- and E-cadherin

based AJs 321, 334. It was therefore possible that Fps might serve a similar function in the

mammary gland, and this might also account for mammary gland dysfunction or a

potential tumor suppressor function in the breast 55. We therefore examined the

composition of the E-cadherin AJs in lactating breast. There were no differences in the

steady state expression levels of E-cadherin, p120 catenin, β-catenin or α-catenin in

lactating fps-null breast tissue relative to wild-type tissue (Figure 2.8A). When Fps/Fer

IPs from wild-type and fps-null tissue were probed for E-cadherin, a prominent band

corresponding to E-cadherin was observed in wild-type tissue, but was essentially

undetectable in fps-null tissue (Figure 2.8B). A band corresponding to E-cadherin was

observed in fps-null tissue in longer exposures. Similar amounts of E-cadherin were

detected in wild-type and fps-null breast by IP followed by immunoblotting. Therefore, in

the wild-type lactating mammary gland, Fps is strongly associated with E-cadherin.

Furtheremore, considering the relative levels of Fps and Fer that were found to be

expressed in the lactating gland (Figure 2.1), we would also speculate that, in epithelial

cells of the lactating breast, E-cadherin interacts to a greater extent with Fps than it does

with Fer.

To determine if the presence of Fps in the AJ was developmentally regulated,

tissue from different stages of pregnancy and lactation were analyzed in a similar manner.

Fps/Fer IPs probed for E-cadherin showed an intense E-cadherin band during stages of

86

87

Figure 2.8 The Fps tyrosine kinase is a component of the E-cadherin-based adherens junction in the mammary gland during lactation. (A) Soluble cell lysates of lactating breast tissue from wild-type (WT) or fps-null (N) mice were immunoblotted (IB) with the indicated antibodies specific for Fps/Fer, E-cadherin, β-catenin, p120-catenin and tubulin. (B) Fps/Fer and E-cadherin IPs of wild-type and fps-null lactating tissues were immunoblotted with an E-cadherin antibody. (C) The presence of Fps in the E-cadherin complex was analyzed during different stages of pregnancy and lactation. Lysates were subjected to Fps/Fer and E-cadherin IPs followed by immunoblot analysis with the reciprocal antibody. (D) Fps/Fer IPs of lactating wild-type and fps-null glands were immunoblotted for E-cadherin, β-catenin and p120-catenin. For controls, Fps/Fer and Fer IPs were probed with the Fps/Fer antibody. (E) Fps/Fer, Fer, E-cadherin, β-catenin and p120-catenin IPs of lactating wild-type and fps-null glands were immunoblotted with the Fps/Fer antibody.

88

lactation (Figure 2.8C upper panel), while faint bands were present during later stages of

pregnancy after longer exposures (data not shown). E-cadherin IPs probed for Fps/Fer

revealed a lactation-specific band corresponding to Fps, and there was also a faint signal

corresponding to Fer (Figure 2.8C lower panel). β-catenin and p120-catenin were also

present in Fps/Fer IPs from wild-type and fps-null samples; and each of these AJ proteins

was more abundant in the Fps/Fer IPs from wild-type tissues (Figure 2.8D). Reciprocal

IP immunoblotting experiments identified both Fps and Fer in IPs of E-cadherin, β-

catenin and p120-catenin from wild-type glands, while only Fer was observed in the IPs

from fps-null tissue (Figure 2.8E). Together these data indicate that both Fps and Fer are

components of the E-cadherin-based AJ in the lactating mammary gland, and may

therefore have roles in regulating cell-cell adhesion and gland function.

2.4.9 Co-localization of the Fps tyrosine kinase with E-cadherin in epithelial cells

during lactation and altered E-cadherin and β-catenin staining in fps-null mammary

glands.

We next examined co-localization between Fps and AJ proteins using

immunofluorescence microscopy. Sections of lactating tissue from wild-type, fps-null

and c-fes mice were probed with α-Fps/Fer and α-E-cadherin antibodies. Although Fps

appeared to be distributed throughout the cytoplasm in the wild-type gland, a fraction of

it was found to co-localize with E-cadherin along the basolateral surface of the epithelial

cells. This was more apparent in the c-fes glands which express several fold more human

Fps than endogenous mouse Fps protein (Figure 2.9A). The E-cadherin staining pattern in

89

Figure 2.9 Co-localization of the Fps tyrosine kinase with E-cadherin in the lactating mammary gland and the subcellular localization of E-cadherin and β-catenin in glands from fps-null mice. (A) Fps/Fer (red) and E-cadherin (green) were visualized in c-fes (upper), wild-type (middle) and fps-null (lower) glands during lactation. A distinct co-localization signal (yellow) of Fps and E-cadherin was observed in the c-fes and wild-type overlay images. (B) Wild-type and fps-null samples were stained with Fps/Fer (red) and E-cadherin or β-catenin (green) to visualize differences in their subcellular localizations between the genotypes.

90

the fps-null tissues was similarly concentrated at the basolateral surfaces, but appeared

less organized than in wild-type tissues. Analysis of Fps and β-catenin revealed a similar

pattern of co-localization at basolateral surfaces, and a relatively less organized β-catenin

localization in fps-null tissues (Figure 2.9B).

2.4.10 Protein-protein interactions between the components of the adherens junction.

The altered localization of E-cadherin and β-catenin in lactating fps-null breast

tissue suggested that Fps might regulate interactions between the different AJ proteins.

To investigate this possibility, AJ components were IPed from wild-type and fps-null

lysates, and the amount of other associated AJ components were assessed by

immunoblotting. Although the interaction between E-cadherin and β-catenin was the

same in wild-type and fps-null samples, there was a reduction in the amount of E-

cadherin and β-catenin that interacted with p120-catenin in the fps-null tissue (Figure

2.10).

2.4.11 Endogenous phosphotyrosine levels of AJ components.

The presence of Fps and Fer in the AJ suggested that one or more of the

components was likely to be a kinase substrate. The in vivo phosphotyrosine levels of E-

cadherin, β-catenin and p120-catenin were compared between wild-type and fps-null in

lactating breast tissue. There was no detectable in vivo tyrosine phosphorylation of any of

these AJ proteins analyzed from either genotype (Figure 2.11). Even in the Fps/Fer IPs

91

Figure 2.10 Differences in the interactions between the adherens junction components from wild-type and fps-null mammary glands. Soluble cell lysates of lactating wild-type and fps-null glands were IPed for E-cadherin, β-catenin and p120-catenin and then immunoblotted with specific antibodies as indicated.

92

Figure 2.11 Endogenous phosphotyrosine content of the adherens junction components. Fps/Fer, Fer, E-cadherin, β-catenin and p120-catenin were IPed from lysates of wild-type and fps-null glands during lactation and immunoblotted with α-pY99 to observe their tyrosine phosphorylation (upper panel). Control immunoblots were performed to determine the amount of protein present in each IP (lower panel).

93

from wild-type tissue, except for phosphorylated Fps, there were no other

phosphotyrosine bands observed.

2.4.12 Fps and Fer kinase activities are associated with p120-catenin.

It was possible that our inability to detect differences in the in vivo

phosphorylation status of wild-type and fps-null AJ components could be due to

relatively low levels of phosphorylation during lactation. Phosphorylation in AJs has

been reported to be reduced in tightly packed cells 370. We have determined that Fps itself

is highly phosphorylated in vivo in lactating breast tissue (Figures 2.2 and 2.11) and that

it is associated with AJ components (Figures 2.8 and 2.9). We therefore performed

immune-complex kinase assays to determine of Fps or Fer could phosphorylate co-IPed

AJ components. IPs of Fps/Fer, Fer, E-cadherin, β-catenin, α-catenin, p120-catenin were

incubated in the presence or absence of ATP and then subjected to anti-phosphotyrosine

immunoblotting analysis. As expected, the appropriate sized phosphotyrosine bands for

Fps or Fer were present in the Fps/Fer and Fer IPs after incubating with ATP (Figure

2.12). Surprisingly, although we have shown that Fps and Fer co-immunoprecipitate with

E-cadherin and β-catenin, phosphotyrosine bands were not reproducibly observed in

these immune complex kinase assays. However, phosphotyrosine bands corresponding to

the size of Fps and Fer were observed in the p120-catenin immune complex kinase assays

from wild-type tissues but not fps-null tissues.

94

Figure 2.12 In vitro kinase assays of the adherens junction components. Fps/Fer, Fer, E-cadherin, β-catenin, α-catenin and p120-catenin were immunoprecipitated from lysates of wild-type and fps-null glands during lactation and incubated in the absence or presence of ATP. Tyrosine phosphorylation was detected by immunoblotting with α-pY99.

95

2.5 Discussion

A previous report that Fps mRNA levels were differentially regulated during

pregnancy and lactation suggested that the Fps kinase may regulate some aspect of

mammary development or function 107. More recently it was suggested that Fps may

function as a suppressor of breast tumorigenesis, as either the loss or inactivation of Fps

resulted in accelerated tumor growth 55. Together, these data indicate that Fps is an

important regulator of the mammary gland. However, they do not provide insights into

the molecular function of Fps. In this report, we further the understanding of the role for

Fps within the mammary gland and identify some important interacting molecules.

Because mRNA and protein expression do not always correlate 371 it was

important to determine if the RT-PCR data would translate to the level of the Fps protein.

In agreement with the Chodosh report 107 we observed a concomitant increase in Fps

protein levels as pregnancy progressed. This culminated at parturition, where Fps

expression reached maximal levels and remained high throughout lactation. Based on its

level of tyrosine phosphorylation, Fps was also highly activated during lactation. This

result was quite surprising as the activation of Fps is known to be tightly regulated 86, 87,

349, 351 and is typically found in an unphosphorylated/inactive state in both tissues and cell

lines 349. The parallel between the increase in Fps expression and the progression of

pregnancy indicated that Fps may be involved in mammary development, possibly as a

positive regulator of cell proliferation, survival or differentiation. Several studies have

shown that activation of Fps in myeloid progenitor cells promoted their survival and

terminal differentiation into macrophage or granulocytic cells 94, 353-357, 372. Similarly, Fps

96

has been linked to neuronal outgrowth and differentiation 348. This led us to believe that

Fps could be involved in epithelial expansion and maturation during pregnancy.

However, its highest expression and robust activation after parturition suggests

that Fps may have a more prominent role in promoting the survival of the fully

differentiated epithelial cells or regulating some aspect of milk production and secretion.

Consistent with this premise was the observation that pups reared by fps-null females

showed a slower rate of weight gain than wild-type counterparts. These results are

suggestive of a lactation defect in the fps-null mammary gland. If lactation is

compromised in the fps-null mice, then the milk should have qualitative or quantitative

differences compared to milk from wild-type mice. Surprisingly, no differences in the

milk proteins (either total protein or specific proteins) or triglycerides were observed. The

total amount of milk produced over a given time was not determined. So, even though

there were no qualitative differences between the two genotypes we could not rule out a

role for Fps in regulating the rate of milk production. The concentration of milk lactose

was also not quantified. The proper concentration of lactose is critical to milk quality and

quantity. If the level of lactose is altered in fps-null milk then it could provide a possible

explanation for the differences in the observed pup weights. However, a reduction in milk

volume and an increase in protein and fat concentration are observed in the absence of

lactose 373. Because no substantial differences in fat and protein levels were observed in

fps-null milk it is unlikely that lactose levels would be altered.

Previously, Fps has been shown to co-localize with the trans Golgi network

proteins, γ-adaptin and TGN38, and the Rab GTP-binding proteins involved in

97

endocytosis (Rab5B and Rab7) or exocytosis (Rab1A and Rab3A) 52, 63, which is

suggestive of a possible role for Fps in vesicle trafficking. Rab3A is known to be

expressed in mammary gland epithelial cells, though its expression is not up-regulated

during lactation 374. Increased expression of six other Rab proteins during lactation has

been reported 219, suggesting that some Rab proteins are important to secretory events

involved in milk release. Currently, it is not known if Fps can regulate the function of any

Rab proteins.

The expression pattern of Fps is known to be restricted to certain types of cell. In

addition to cells of the hematopoietic system, Fps is expressed in endothelial, neuronal

and some epithelial cells 51, 52, 348. However, the cell-specific expression pattern of Fps in

the mammary gland had not been addressed. As the increase in Fps expression coincided

with massive epithelial cell proliferation during pregnancy it was likely that Fps would be

found in epithelial cells. By using antibodies that recognize Fps, an epithelial cell-

specific marker, Mucin-1 375, and a myoepithelial marker, smooth muscle actin (SMA)

369, we have identified epithelial cells as the primary source of Fps expression in the

mammary gland. Fps mRNA has been reported in Sp1 cells, a murine metastatic

mammary tumor cell line 52. But to the best of our knowledge this is the first

documentation of Fps expression in normal mammary epithelial cells. There was no co-

cellular expression between Fps and SMA indicating that Fps is not found in

myoepithelial cells. Other cells, like endothelial cells, are known to express Fps but they

are not thought to contribute a significant amount to the overall level of Fps expression

found in the mammary gland during lactation.

98

To attempt to assess the biochemical role of Fps during lactation we examined the

phosphotyrosine status of several potential Fps substrates. Fps can interact with and

phosphorylate several Stat family members, including Stat1, Stat3 and Stat5 22, 376, 377.

The importance of Stat5 to mammary development and expression of specific milk

proteins is well established. Stat5 is known to be critical to epithelial proliferation,

differentiation and survival during pregnancy and also to the maintenance of epithelial

differentiation and survival during lactation 148, 378, 379. However, the tyrosine

phosphorylation of Stat5 during lactation was not reduced in the fps-null tissue. Further

evidence that suggested Fps was not a component of the prolactin-Stat5 pathway was

provided by whole-mount analysis that showed there was no obvious difference in the

number or morphology of the alveolar structures during pregnancy or lactation between

the two genotypes.

The report that examined Fps expression in the mammary gland also showed that

Akt1 was maximally expressed during lactation 107. Akt1 is involved in lipid synthesis 165,

166, glucose metabolism 165 and cell survival 380, 381 during lactation. Fps has been shown

to activate Akt in vascular endothelial growth factor-A stimulated endothelial cells 382,

and has been implicated in the activation of PI3K, an upstream regulator of Akt, in both

endothelial and neuronal cells 348, 383-386. The fact that the expression patterns of Fps and

Akt1 were markedly similar suggested that they may be components of a common

pathway during lactation. However, when Akt and p85, the regulatory subunit of PI3K,

were analyzed, there were no differences in their level of phosphorylation between the

99

genotypes. This suggests that Fps is unlikely to participate in PI3K-Akt regulated

pathway(s) during lactation, or to have a role in glucose metabolism or lipid synthesis.

We also examined the tyrosine phosphorylation status of other proteins involved

in different aspects of exocytosis and endocytosis. Proteins such as N-ethylmaleimide-

sensitive fusion protein (NSF), Dynamin II and Pacsin-2 from wild-type and fps-null

glands displayed no phosphotyrosine differences. Analysis of the signaling molecules

Erk, FAK, cortactin, p130Cas, and p190RhoGAP, which are potential substrates or

downstream effectors of Fps, also showed no differences.

Although there were no obvious morphological differences that could be

ascertained from the wild-type and fps-null whole-mounted mammary glands, this

analysis was not useful to examine the alveolar microstructure. Thus, thin sections of

each tissue were H&E stained and compared at the cellular level. There were no obvious

differences in the size or number of alveoli, which provided further evidence that the

development of the fps-null glands is not compromised. Higher magnification images

showed that within individual alveoli, the lateral intercellular contacts between the

epithelial cells were not significantly different between the genotypes. These data suggest

that Fps does not have an important role in mammary gland development but may

possibly regulate epithelial cells in terms of their ability to function at full capacity during

lactation.

In our quest to identify interacting partners and substrates of Fps, candidates were

chosen whose altered function may promote tumorigenesis and result in abnormal

mammary function. Components of the E-cadherin-based adherens junction complex

100

were analyzed based on these criteria. Also, Fer is a known component of both E- and N-

cadherin based AJs 334, 387 and some functional redundancy has been shown to exist

between Fps and Fer 49. Thus, it was possible that in types of cells where Fps is much

more highly expressed, like lactating epithelial cells, it could replace or compliment the

function of Fer. Our initial experiments revealed that E-cadherin was indeed associated,

either directly or indirectly, with Fps, and to a lesser extent with Fer. This result was

particularly intriguing as it suggested that Fps may have a role in regulating cell-cell

adhesion.

By analyzing tissues from different developmental stages we next determined that

Fps and E-cadherin were found in a complex predominantly during lactation and to a

lesser extent during late stages of pregnancy. The connection between Fps and E-

cadherin was supported further by confocal microscopy where a distinct signal of co-

localization was observed. The presence of Fps in the AJ during lactation suggests that it

may provide an important function to the AJ that is specifically required during this

stage. At the onset of lactation most of the epithelial cells have stopped proliferating and

have become fully differentiated 144. So, even though a low proportion of the AJ

components are likely to be in a state of continual turn-over, the AJs are considered

mature and stable during lactation. It is possible that Fps gets recruited to the AJ after the

initial cell-cell contacts are established and the AJ has formed a stable structure.

When the analysis was extended to β-catenin and p120-catenin, they too were

found to be in a complex with Fps and Fer during lactation. It did not appear that Fps

preferentially associated with one of the components over another which may be

101

indicative of an indirect interaction or a weak affinity for its interacting partner(s). Fer is

known to associate directly with p120-catenin 321, 367. This interaction was mapped to a

thirty amino acid sequence (331-360) in the second predicted coiled-coil motif of Fer 321.

There is less than twenty percent homology between Fps and Fer in this region. Although

this does not rule out a Fps/p120-catenin interaction it could mean that Fps binds to

another AJ component.

Nevertheless, because Fps and Fer are known to share some functional

redundancy, it is possible that Fps serves a similar purpose in the AJ as Fer. Much more

is known about the role of Fer in regulating the phosphorylation of AJ components. Its

over-expression or loss of function has dramatic effects on cell-cell contacts in E-

cadherin and N-cadherin expressing cells, respectively 321, 334. Fer over-expression

promotes its association with and phosphorylation of both p120-catenin and β-catenin.

This leads to a decrease in the amount of α- and β-catenin associated with E-cadherin and

a subsequent loss of intercellular adhesion 334. Interestingly, the absence of Fer results in

the mislocalization of N-cadherin and β-catenin from fibroblast cell-cell contacts and the

weakening, or loss, of intercellular adhesion 321, 388. In these cells, one of the biochemical

functions of Fer is to phosphorylate the tyrosine phosphatase, PTP1B, on Y152 321 which

promotes its association with N-cadherin 333. PTP1B then maintains the

dephosphorylation of Y654 of β-catenin and prevents its dissociation from the cadherin

complex 321. In this context it appears that the normal function of Fer is to strengthen

cadherin-based cell-cell adhesion, but too much or too little Fer ultimately has a negative

effect.

102

It is interesting that both Fps and Fer were detected in the AJ. Previous to this

study, only Fer was believed to have a role in regulating AJ components. The presence of

Fer in the AJ of these cells suggests that its traditionally accepted function could be

preserved. However, the differences in the level of expression between Fer and E-

cadherin/β-catenin indicate that either Fer is not required in a 1:1 ratio with these proteins

or additional kinases are present in the AJ. It is well established that other kinases such as

members of the Src-kinase family are present in AJs and are important regulators of

different AJ components. Src can phosphorylate p120-catenin under certain conditions

335. p120-catenin can act as a docking protein for the recruitment of Yes and Fyn to the

AJ and facilitates their activation 320. The activation of Fyn can directly lead to β-catenin

Y142 phosphorylation and the dissociation of β-catenin from α-catenin 320. Thus, the

presence of another kinase, like Fps, in the AJ should not be overly surprising.

Despite the substantially higher levels of Fps when compared to Fer, it is difficult

to believe that Fps substitutes for Fer in the AJ during lactation. If this were true, then the

phenotype within the fps-null mammary gland should display some similarities to that

observed in Fer-mutant fibroblasts 321. Specifically, unphosphorylated PTP1B would not

be able to interact with E-cadherin, which would lead totyrosine phosphorylation of β-

catenin and its accumulation in the cytoplasm as it dissociates from the AJ. Clearly the

data presented here do not reflect this scenario. In fact, β-catenin displayed no

phosphorylation on tyrosine residues in wild-type or fps-null tissue, and was strongly

associated with E-cadherin regardless of the genotype, meaning the mechanisms

regulating its tyrosine phosphorylation are intact in the fps-null tissue. Tyrosine

103

phosphorylation also was not detected for either E-cadherin or p120-catenin. This is not

considered unusual because tyrosine phosphorylation of E-cadherin is not thought to be a

prominent event in normal stable tissues 285. Although tyrosine phosphorylation of p120-

catenin has been documented in response to multiple stimuli 317, it is believed to be

greatly reduced in confluent cells 370 and likely in tissue as well. The data suggest that

during lactation E-cadherin, β-catenin and p120-catenin are not substrates of Fps or Fer

or other tyrosine kinases whose activity is counterbalanced by a Fps-dependent tyrosine

phosphatase.

Our results indicate that the loss of Fps does not affect the interaction between E-

cadherin and β-catenin. This is supported by the equal amounts of E-cadherin and β-

catenin observed in the reciprocal IPs in wild-type and fps-null tissue. In contrast to these

results is the apparent reduction in the amount of E-cadherin and β-catenin that was

present in p120-catenin IPs of fps-null tissue. Despite its reproducibility, there was not

less p120-catenin pulled down in the IPs of either E-cadherin or β-catenin from fps-null

tissue. This is perplexing and currently without a solid explanation. Typically, the

uncoupling of p120-catenin from E-cadherin destabilizes the E-cadherin/catenin complex

and results in its internalization with endocytic vesicles 255. Once internalized, the

complex can be recycled to the surface if sufficient free p120-catenin is available 254, or it

can be targeted for proteolytic degradation through ubiquitin-dependent or -independent

processes 285, 389. p120-catenin is known to have a low affinity for cadherins 252 which

may be a reflection of its ability to rapidly alternate between cadherin-bound and -

unbound states. This characteristic also may contribute to the dynamic nature and

104

continuous turnover of AJ components. Our results suggest that the loss of Fps causes a

fraction of the total amount of E-cadherin-bound p120-catenin to become uncoupled from

the complex. Theoretically, the likely effect of this would be an increase in the

cytoplasmic pool of p120-catenin and a destabilization of E-cadherin that would get

internalized. In support of this is the differences observed in the immunofluorescent

staining patterns of E-cadherin and β-catenin in the wild-type and fps-null tissues. In the

wild-type samples, both E-cadherin and β-catenin displayed a clear localization to the

lateral and basal surfaces of the epithelial cells. In the fps-null tissue, E-cadherin and β-

catenin were also found at these surfaces but the staining appeared less distinct and less

localized to the cell-cell junctions than in the wild-type tissues. Although an increase in

cytoplasmic E-cadherin or β-catenin was not readily apparent, the overall pattern of

staining was more disorganized in the fps-null tissues.

How the loss of Fps would cause the dissociation of p120-catenin from E-

cadherin is currently unknown. The low affinity of p120-catenin for E-cadherin could

make it susceptible to aberrant uncoupling from E-cadherin in the absence of Fps if it

comes into contact with a higher affinity binding partner. p120-catenin is known to

interact with multiple proteins, both within and outside the AJ 261. For instance, p120-

catenin is known to interact with activated (tyrosine phosphorylated) p190RhoGAP at the

AJ, which then down-regulates RhoA GTPase activity 390. Despite the translocation of

p190RhoGAP to the AJ, it interacts strongly only with p120-catenin but not other

members of the cadherin complex 390. This suggests that interactions with non-AJ

components may be sufficient to uncouple p120-catenin from the AJ. Regardless of the

105

biochemical function of Fps in the AJ, it is well established that the proper expression

and normal function of the AJ components is critical to epithelial cell shape, polarity,

differentiation and survival (reviewed in 360). If some aspect of AJ function is altered,

then it could conceivably contribute to the suspected lactation defect observed in the fps-

null mice.

In summary, our findings indicate that Fps regulates mammary gland function

during lactation. The presence of Fps in the E-cadherin-based AJ specifically during

lactation suggests that it may contribute to an aspect of cell-cell adhesion that is

important to the full functional capacity of the gland. Further studies will be needed to

elucidate the precise function of Fps in regulating the AJ and how its loss may negatively

impact lactation.

106

Chapter 3

Immortalization of murine mammary epithelial cells by targeting p53 expression

with a lentivirus-mediated RNA interference system.

3.1 Abstract

The Fps tyrosine kinase is expressed in murine mammary epithelial cells.

Analysis of gene targeted and transgenic mice suggests that Fps functions as a tumor

suppressor in the mammary gland and regulates lactation. To better understand the

function of Fps in the mammary gland our initial studies focused on the analysis of

primary cells from wild-type and fps-null tissues. With these cells we were able to induce

Fps expression after treatment with lactogenic hormones. Subsequent to these studies we

have developed a system to establish immortalized epithelial cells from fps-null mice.

Using established methods, primary epithelial cell organoids were initially isolated from

mid-pregnant mice and then enriched in culture. The highly purified epithelial cell

population was then infected with a lentivirus expressing a mouse-specific p53 shRNA.

A separate GFP reporter gene in the virus was used to assess viral transduction and,

indirectly, the “knockdown” of p53 expression. In order to stimulate growth and maintain

the epithelial phenotype the population of infected and uninfected cells was grown on

collagen gels for several passages. By monitoring GFP expression there was an obvious

enrichment of infected cells and a reduction of uninfected cells. Immunofluorescence

analysis of early passage epithelial cells with antibodies to E-cadherin and β-catenin

showed that their expression was maintained and they localized to cell-cell junctions.

107

With additional rounds of growth an immortalized cell line was established. Immunoblot

analysis demonstrated that these cells express a combination of epithelial and non-

epithelial cell markers. Immunofluorescence analysis of the immortalized cells revealed

that E-cadherin was internalized in vesicular structures, whereas N-cadherin and β-

catenin were observed at the cell periphery. Based on these data, it is likely that the fps-

null epithelial cells have acquired characteristics associated with an epithelial to

mesenchymal transition. Their utility as a tool to investigate the normal function of Fps in

the mammary gland is limited. However, they may be of use in determining the role of

Fps in mammary tumorigenesis.

3.2 Introduction

Primary cells represent one of the most suitable model systems for biochemical

and signal transduction studies in a controlled setting. However, primary cells are

restricted by their limited proliferative potential in culture and can vary in their

homogeneity between cell preparations 391. The limited growth of cells in culture can be

accounted for in part by an increased susceptibility to apoptosis and an induction of

senescence 392. Thus, if the conditions or factors that cause these two processes can be

avoided, then primary cells could divide in culture to a point were they become

immortalized, thereby establishing cell lines.

One of the key proteins regulating the induction of both apoptosis and senescence

is the p53 tumor suppressor protein 393, 394. It is well established that p53 is lost or

functions incorrectly in many types of cancer 395, and that it plays a crucial role in the

108

prevention of tumor development by causing growth arrest or apoptosis of abnormal cells

396. In vivo, the p53 signaling pathway is inactive under normal cellular conditions, but

becomes activated in response to cellular stress signals that arise from such things as

DNA damage, abnormal proliferation, loss of cell-cell adhesion, hypoxia and telomere

shortening 397, all of which have been associated with tumorigenesis 398-400. Each stress

causes the activation of one or more signaling pathway which, in turn, leads to the post-

translational modification and activation of p53 401. Primarily, p53 activation, involving

an increase in both its stabilization and DNA binding ability 402, is controlled by

phosphorylation 403 and acetylation 404. It is believed that the combination of post-

transcriptional events dictates the level of p53 activation which in turn activates the

appropriate cellular response to the detected stress 401. Although p53 possesses functions

separate from transcriptional regulation, its primary role is as an activator of

transcription. The genes regulated by p53 can be divided into four general categories; 1)

cell cycle inhibitors 405; 2) DNA repair 406; 3) apoptosis regulators 407; and 4) inhibitors of

metastasis and angiogenesis 408, 409. Typically, the end result is the temporarily or

permanent arrest of cell division, or self-removal by apoptosis if the damage is severe.

The growth arrest or death which is experienced by primary cells in culture is

caused by stresses that are different from those found in vivo. It is thought that these

“tissue culture stresses” are the result of an inadequate culture environment which can

involve an improper extra-cellular matrix, excessive bovine growth factors and non-

physiological oxygen concentrations 410. Nevertheless, the stress pathways activated by

these culture conditions converge to activate p53, resulting in similar impaired growth or

109

death as that observed in in vivo systems 410. Through either mutation or deletion, the

function of p53 is often disrupted in both immortalized and transformed cell lines 411.

Because p53 is an important regulator of cell growth, and its activation is likely to occur

in primary cells, targeting its expression is likely to facilitate the transition from the

primary cell stage to the immortalized stage 411, 412.

The use of RNA interference has become a useful tool to silence gene expression

and study protein function 413, 414. It was anticipated that the “knockdown” of p53

expression would be a useful approach to allow primary cells to become insensitive to the

stress signals that lead to senescence or apoptosis. And, it would enhance our ability to

establish immortalized cell lines without using a viral or cellular oncogene. To

accomplish this, we elected to use a lentivirus based system to express a p53 mouse-

specific short hairpin (sh) RNA 415. There are several advantages with the lentiviral

system, including the ability to infect non-dividing cells, the stable integration into the

host genome and the sustained high-level transgene expression. This type of system has

been used extensively to silence many different genes (reviewed in 416).

With this method we chose to immortalize mammary epithelial cells from fps-null

mice. The Fps/Fes protein tyrosine kinase was originally thought to be restricted to cells

of the hematopoietic system, and numerous studies have shown that it is involved in

several different signaling pathways in the hematopoietic system (reviewed in 5). Fps

expression has also been observed in endothelial, neuronal and some epithelial cells 52.

fps transcripts have been detected in the mammary gland 107. We have confirmed and

extended this observation at the protein level showing that Fps is highly induced during

110

lactation and is expressed in epithelial cells of the mammary gland (Truesdell and Greer,

unpublished). Additional studies have shown that in a MMTV/PymT breast cancer mouse

model the absence of the Fps protein caused tumors to develop with an earlier onset than

in the presence of Fps, suggesting that Fps may function as a tumor suppressor in breast

epithelial cells 55. However, because Fps regulates the innate immune system 26, 102, 103

and the vascular system 51, 52, 99 both of which can influence tumorigenesis, it is unknown

whether the absence of Fps in cell types other than breast epithelial cells might contribute

to this apparent tumor suppressor effect.

To further characterize the function of Fps tyrosine kinase in the mammary gland

we have established a procedure whereby primary fps-null mammary epithelial cells were

infected with a lentivirus encoding p53 shRNA and allowed to proliferate on a collagen

gel substratum for several passages. From the mixture of primary cells, transduced

epithelial cells continued to proliferate as nontransduced cells declined in number. After

several passages, only the transduced cells remained. As the mixture of cells was

continually expanded, clonal cell populations emerged that appeared epithelial in nature.

These cells were isolated, expanded and analyzed with respect to their epithelial

characteristics.

3.3 Materials and Methods

3.3.1 Materials

Rat tail collagen, collagenase A and BSA fraction V were purchased from Roche

Applied Science. Insulin, DMEM/F12, hyaluronidase, nystatin, DMEM, linoleic acid,

111

dexamethasone and prolactin were purchased from Sigma-Aldrich. Human EGF was

from R&D Systems and FBS was from Hyclone. The antibiotics/antimycotic solution and

RPMI-1640 were obtained from Invitrogen. The 70 µm cell strainers were supplied by

BD Biosciences.

3.3.2 Antibodies

The anti-Fps/Fer antibody was generated as described previously 52. The primary

antibodies (company) were: E-cadherin, N-cadherin, β-catenin (BD Biosciences); SMA

(Sigma-Aldrich); p53 and EGFR (Santa Cruz Biotechnology); Vimentin (gift from Dr.

Bonnie Asch). The secondary antibodies were: Alexa Fluor 488- and 633-conjugated IgG

F(ab)2 fragments (Molecular Probes); goat anti-rabbit horseradish peroxidase IgG (Vector

Laboratories) and goat anti-mouse horseradish peroxidase IgG (GE Healthcare).

3.3.3 Lactogenic hormone induction

Primary mammary epithelial cells from both fps-null and wild-type mice were

grown to confluency on collagen-coated surfaces. The cells were washed in PBS and then

grown in DMEM/F12 with 10 % FBS, dexamethasone (1.0 µM), insulin (5 µg/ml), and

prolactin (5 µg/ml) for 4 days.

3.3.4 Generation of p53 RNAi lentivirus and MSCVpac/Fpsmyc virus

The Cla I/Eco RI fragment from the pSuper-p53 shRNA plasmid (kindly provided

by Rene Bernards) containing the H1 promoter and murine p53 shRNA sequence was

112

cloned into the lentivirus vector pLVTHM using the same restriction sites. The resulting

vector plasmid, pLVTHM-shp53, along with the pCMVΔR8.91 packaging plasmid, and

the pMD.G envelope plasmid, were used to produce non-replicating lentivirus following

transfection of 293T packaging cells. The 293T cells were transfected with the three

plasmids using a calcium phosphate transfection method developed by the Didier Trono

Lab (http://tronolab.epfl.ch). The lentiviral medium was collected at 48 and 72 hrs post-

transfection, combined, filtered through a 45 µm filter and then used for infection or

frozen at -80oC. Infection of epithelial cells involved overlaying the cells with the virus

medium and incubating overnight at 37oC.

The MSCVpac/Fpsmyc virus was produced in 293T cells which were transfected

with MSCVpac/Fpsmyc and EcoPac plasmids using a standard calcium phosphate

transfection protocol. After 48 hours the supernatant was collected, filtered,

supplemented with 10 µg/ml Insulin, 5 ηg/ml EGF and 5 µg/ml polybrene, and then

added to the cells for 24 hrs. Puromycin selection (2 µg/ml) was performed over a 7 day

period.

3.3.5 Transgenic mice and isolation of epithelial cells

The generation of in-bred SVJ/129-CD1 hybrid wild-type and fps-null mice has

been described previously 26. Female mice were paired with males for mating and

vaginal plug dates were recorded. At day 12-14 of pregnancy, mammary glands were

aseptically removed and tranferred to a culture dish. The digestion of tissue and epithelial

growth was adapted from 446. The tissue was minced with scissors into a fine paste and

113

then added to digestion medium (DMEM/F12, 1 % antibiotics/antimycotics, 60 U/ml

nystatin, 2 mg/ml collagenase A and 100 U/ml hyaluronidase) at 1 g of tissue per 10 ml

medium. The homogenate was incubated in a 37oC water bath shaker at 120 rpm for 2 hrs

with gentle pipetting every 30 minutes to ensure complete digestion. After filtration

through a 70 µm cell strainer the epithelial cells and organoids were washed extensively

with 5 % FBS in PBS. The cells were then resuspended in a neutralized collagen solution

(1.7 mg/ml collagen, 0.02 N NaOH, 1 x DMEM), plated onto a pre-solidified collagen

gel and allowed to solidify for 1 hr. The imbedded cells were overlaid with epithelial

medium (DMEM/F12, 1 mg/ml BSA Fraction V, 5 µg/ml linoleic acid, 10 µg/ml insulin,

5 ηg/ml EGF, 1 % antibiotics/antimycotics, 20 U/ml nystatin) and grown for 1 to 2 weeks

until the gel contracted away from the plate perimeter. To digest the gel, collagenase A

was added to the growth medium at 2 mg/ml and incubated at 37oC until all the gel was

solubilized. The cells were washed in 5 % FBS in PBS as before, resuspended in growth

medium and then allowed to adhere to a pre-solidified gel and recover for 24 hrs. The

majority of the cells were in small clusters of approximately 40-100 cells which would

adhere and then spread out onto the gel surface.

Supernatant containing the pLVTHM lentivirus encoding mouse-specific p53

shRNA was supplemented with 10 µg/ml Insulin, 5 ηg/ml EGF and then added to the

cells for three consecutive days, with fresh virus added each day. Plates were grown for

three additional days before assessing infection efficiency based on GFP expression. The

mixture of cells was grown until the cells had expanded to approximately 70-80 % of the

plate. To passage the cells, the plate was treated with 5 mM EDTA in PBS at 37oC until

114

cell-cell contacts began to dissociate. The gel was then digested with collagenase A and

washed as before. Usually the cells were re-plated onto new collagen gels at 1/2 to 1/3 of

their original density for additional growth. This cycle was repeated approximately eight

to ten times until a population of cells that retained an epithelial morphology eventually

became the prominent cell type. Occasionally during this process small groups of cells

that possessed a non-epithelial morphology would begin to grow. The location of these

cells was marked and the cells were gently scraped off the surface of the gel with a pipet

tip and removed by aspiration. After it appeared that an immortalized cell population was

present the cells were transferred to collagen coated culture plates for expansion and

analysis. Plates were coated with 50 µg/ml collagen for 1 hr at 37oC, washed with PBS

and then used or stored at 4oC for up to 1 week. At this stage the epithelial medium was

changed where 10 % FBS was used in place of 1 mg/ml BSA.

HC11 cells were grown in RPMI-1640 supplemented with 10 % FBS, 5 µg/ml

insulin and 10 ηg/ml EGF. Cells were infected with the p53 shRNA virus with the same

protocol used for primary cell infection.

3.3.6 Immunofluorescence

Cells grown on collagen coated coverslips were fixed with 3 % paraformaldehyde

in PBS for 15 minutes and then permeabilized with 0.2 % Triton-X 100 in PBS. After

blocking in 3 % BSA in PBS for 30 minutes, the cells were overlayed with primary

antibody and incubated at RT for 1 hr. All primary antibodies were diluted 1:100 in

blocking buffer. The cells were then washed 3 times in blocking buffer with 0.01 %

115

Triton-X, followed by incubating with the appropriate Alexa fluor-conjugated secondary

IgG F(ab)2 fragment (1:200 dilution in blocking buffer) for 1 hr at RT. Nuclei were

stained with 2 mM Hoechst 33258 for 10 minutes in PBS. Cells were washed 3 times for

10 minutes each with PBS and then mounted with Mowiol (Calbiochem). Images were

captured on a Leica SP2 true scanning confocal microscope.

3.3.7 Western analysis

Cells were lysed in 4oC RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl,

1mM EDTA, 1mM NaF, 1 % NP-40, 1 % Na-deoxycholate, 0.1 % SDS, 10 µg/ml

aprotinin, 10 µg/ml leupeptin, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl

fluoride) for 30 minutes. Whole cell lysates were mixed with 1/5th volume 6x SDS

sample buffer (70 % Tris/SDS stacking gel buffer, 30 % (v/v) glycerol, 10 % (w/v) SDS,

DTT, bromophenol blue) and heated to 95oC for 5 minutes. Protein aliquots of 10 µg

were separated by SDS-PAGE and transferred to Immobilon P membranes (Millipore)

using a semidry transfer apparatus (Bio-Rad). Membranes were blocked with 5 %

skimmed milk in TBST buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1 % Tween-

20) for rabbit antibodies and 5 % BSA in TBST for mouse antibodies, and then incubated

overnight at 4oC with the specified antibody. After washing in TBST, membranes were

incubated with a 1:10,000 dilution of secondary antibody in TBST for 1 hour at RT. After

washing with TBST, reactive proteins were detected by using enhanced

chemiluminescence reagents (NEN Life Science Products).

116

3.4 Results

In order to further explore the potential functions of the Fps tyrosine kinase in

murine mammary epithelial cells we sought to establish cell culture models appropriate

for both biochemical and cell biology analyses. We began by establishing conditions for

growth of primary murine mammary epithelial cells in a collagen gel matrix that has been

reported to support epithelial cell survival and expansion, while limiting fibroblastic cell

growth 446. Colonies of cells grown under these conditions displayed tubule-like

morphologies reminiscent of branching ductal structures in the mammary gland (Figure

3.1A). When isolated from these collagen gel matrices and grown on collagen-coated

plates, the resulting cultures displayed typical cuboidal epithelial cell morphologies

characteristic of normal mammary cells, and these cultures were highly enriched for these

cell types (Figure 3.1B).

We next examined the expression of Fps and the related Fer kinase in these

primary cultures. Immunoblotting analysis using an antibody that is cross-reactive for

Fer and Fps revealed that both kinases were expressed at approximately the same levels

in wild-type cultures. As expected, only Fer was detected in cultures from fps-null

mammary tissues (Figure 3.2). Interestingly, Fps expression was significantly up-

regulated after four days of culture in lactogenic hormones (dexamethasone, insulin and

prolactin) (Figure 3.2). This observation is consistent with our in vivo analysis showing

up-regulation of Fps in the mouse breast during pregnancy and lactation, as described in

Chapter 2.

117

Figure 3.1 Growth of primary murine epithelial cells in a collagen gel matrix and on collagen-coated surfaces. Epithelial cell organoids were isolated from mid-pregnant mouse mammary glands and suspended in a collagen solution that solidified to form a gel. After one to two weeks of proliferation the cells had developed into three dimensional structures (A) that were then purified from the collagen gel and grown as primary epithelial cell cultures on collagen-coated surfaces (B). Magnification was 40 x for (A) and 200 x for (B).

118

Figure 3.2 Induction of Fps expression in wild-type epithelial cells in response to treatment with lactogenic hormones. Wild-type and fps-null primary epithelial cell cultures were grown on collagen-coated surfaces in the presence of dexamethasone (1.0 µM), insulin (5 µg/ml), and prolactin (5 µg/ml) for 4 days. Equal amounts (10 µg) of protein lysates from untreated and treated cells were analyzed by immunoblotting with the Fps/Fer antibody.

119

While cultures of highly enriched epithelial cells could be achieved using this

protocol, we found that yields were inconsistent, making it difficult to expand the

cultures up to amounts needed for biochemical studies. To circumvent these problems we

endeavored to produce immortalized mammary epithelial cell lines that would retain

characteristics of their normal in vivo counterparts. Since loss of p53 expression has been

correlated with establishment of immortalized cells 415 we next attempted to immortalize

these mammary epithelial cells by knocking down p53 expression using RNA

interference. Primary cells obtained as described above were grown on collagen gels and

transduced with lentivirus encoding a shRNA p53 sequence (Figure 3.3) The initial

extent of infection, based on GFP expression, was quite low (< 20 %) (Figure 3.4A).

However, as the population of cells proliferated and was expanded onto fresh collagen

gels several times, the GFP-positive cells became the predominant cell type (Figure 3.4B,

C and D). In parallel experiments involving the use of lentivirus which lacked the shRNA

p53 sequence, cells expressed GFP after infection but quickly died after a couple of

passages (data not shown).

During their expansion and enrichment, some of the transduced cells were

transferred to collagen-coated coverslips and tested for expression of epithelial cell

markers. The cells were positive for both E-cadherin and β-catenin, both of which

showed concentrated localization at cell-cell boundaries, consistent with their presence

within adherens junctions (Figure 3.5, panels A and B). In contrast, there was no specific

staining for the myoepithelial cell marker, smooth muscle actin (SMA) (Figure 3.5, panel

120

Figure 3.3 Schematic representation of the lentiviral p53 shRNA construct used to target p53 expression. The p53 shRNA sequence, along with the H1 promoter, was excised from pRETRO-SUPER-p53 and ligated into the pLVTHM lentivirus vector as an EcoRI/ClaI fragment. The p53 shRNA sequence is shown as a p53-loop-p53. The expression of GFP, shown in green, is controlled by the elongation factor-1 alpha (EF-1 alpha) promoter. Other components to the integration construct include the long terminal repeat (LTR), the central polypurine tract (cPPT) and the woodchuck promoter response element (WPRE).

121

Figure 3.4 Enhanced survival of primary mammary epithelial cells infected with the p53 shRNA lentivirus. Enriched primary epithelial cells grown on a collagen gel layer were infected with the p53 shRNA lentivirus and then monitored for GFP expression. After the initial viral infection a relatively low number of GFP-positive cells were observed (A). After continual growth and passaging on a collagen gel layer, the percentage of GFP-positive cells increased (B and C) until all of the cells were positive for GFP (D). Magnification was 400 x for (A,B,C), and 200 x for (D).

122

Figure 3.5 Retention of epithelial cell properties during the establishment of the fps-null mammary cell line. During the propagation of the p53 shRNA fps-null cells they were examined in terms of their morphology and epithelial marker expression. From (A) and (B) the shape of the cells is cuboidal or polygonal that is characteristic of epithelial cells. When these cells were examined for expression and localization of E-cadherin and β-catenin a distinct red fluorescent signal (Alexa-633) was observed at cell-cell contacts. Cells that had a fibroblast cell shape were examined for SMA expression (C). Nuclei are stain with Hoechst 33258 (blue). Magnification was 400 x.

123

C). These observations suggested that the transduced primary cultures retained epithelial

characteristics.

Although maintaining the cells on a collagen gel served as a good method to

promote epithelial growth, maintain their morphology and limit fibroblast growth, it was

not a convenient way to routinely grow cells in sufficient quantities for biochemical

analysis. Thus, as the cells expanded they were tested for their ability to maintain their

epithelial shape and grow on collagen-coated culture dishes. During early passages on

collagen gels, the shape and growth of the cells was often altered when they were

transferred to collagen-coated surfaces (Figure 3.6A, left panel). However, after

continuous growth on the collagen gel matrix for several (8 to 10) passages, a clonal

population of cells emerged that retained the typical cuboidal epithelial cell shape when

grown on collagen-coated dishes (Figure 3.6B). These cells were isolated and expanded

for analysis.

The expression of specific markers that are associated with either epithelial or

fibroblastic cells were used to evaluate the phenotype of these cells. The fps-null

immortalized cells were compared with lactating breast tissues from wild-type and fps-

null mice, and a normal mouse mammary epithelial cell line, HC11, which is known to

express an inactivated p53 protein 417, or HC11 cells which were infected with the p53

shRNA lentivirus. Immunoblotting analysis with an anti-p53 antibody confirmed that the

transduced fps-null cells lacked p53 expression. For comparison, a p53 signal was

detected in the uninfected HC11 cells but not in HC11 cells transduced with the p53

shRNA lentivirus, or in mammary tissues (Figure 3.7). Additional analysis revealed that

124

Figure 3.6 Growth of epithelial cells at different steps of immortalization. During the establishment of the fps-null cell line the cells were assessed for their ability to grow on collagen coated dishes. The shape and growth characteristics of the cells from early passages would often change when placed on collagen coated dishes. For comparison, a group of cells that retained epithelial properties is shown (A, right panel) along with a group displaying more differentiated morphologies (A, left panel). Eventually a clonal population of cells emerged from the cell mixture. Images of the immortalized cells are shown at low (B, left panel), medium (B, middle panel), and high (B, right panel) densities. Magnification was 200 x.

125

Figure 3.7 Expression of epithelial and non-epithelial markers in the fps-null immortalized cells. Immunoblot analysis was performed on protein lysates from the p53 shRNA immortalized fps-null cells (lane 1), HC11 cells (lane 2), HC11 cells infected with the p53 shRNA lentivirus (lane 3), and wild-type and fps-null mammary glands (lane 4 and 5) using antibodies to p53, E-cadherin, β-catenin, N-cadherin, SMA and EGFR. For the vimentin and Fps/Fer immunoblots, only the fps-null cells (lane 1), HC11 cells (lane 2), and wild-type and fps-null mammary glands (lane 3 and 4) were analyzed.

126

the immortalized fps-null cells retained expression of E-cadherin, but this was

significantly less than the levels seen in HC11 cells and lactating tissues (Figure 3.7).

Similar levels of β-catenin expression were seen in the fps-null and HC11 cells, and these

levels were considerably higher than in the mouse breast tissues (Figure 3.7). The

retention of β-catenin and reduction in E-cadherin expression led us to examine whether

another type of cadherin had been up-regulated during the establishment of these

epithelial cells. A relatively high level of N-cadherin expression was observed in the fps-

null cells. In comparison, HC11 cells expressed a significantly lower level of N-cadherin,

while it was not detected in the breast tissues (Figure 3.7). Expression of vimentin, an

intermediate filament protein commonly found in fibroblast-like cells, was also up-

regulated in the fps-null cells compared to the HC11 and breast tissues (Figure 3.7). In

contrast, very little SMA was detected in the fps-null cells whereas both HC11 and

lactating breast tissues expressed this marker (Figure 3.7). Because the fps-null and HC11

cells depend on EGF for survival we also examined EGFR expression. The EGFR was

absent or below the level of detection in lysates from the lactating tissues. However, the

fps-null cells expressed several fold more EGFR than the HC11 cells (Figure 3.7).

Finally, we assessed the level of Fps and Fer expression in the different cells and tissues.

As expected in the wild-type and fps-null breast tissues, signals were detected

corresponding to Fps and Fer, or just Fer, respectively. The HC11 cells expressed Fer, but

no detectable Fps. Likewise, the fps-null cells expressed only Fer (Figure 3.7).

Due to the apparent reversal in N- and E-cadherin expression in the immortalized

fps-null cells relative to breast tissues, we next investigated the subcellular localization of

127

these cadherins as well as β-catenin. E-cadherin was visualized in cytoplasmic vesicular

structures within the fps-null cells (Figure 3.8A). In contrast, N-cadherin and β-catenin

localized to the periphery of the cells (Figure 3.8B and C), consistent with their being

involved in adherens junctions. During this analysis it was realized that the cells no

longer expressed GFP which had been used as an indicator for the presence of the

lentiviral transgene. However, based upon immunoblotting analysis (Figure 3.7), p53

expression was not detected. Cells were treated with 5-azacytidine, an inhibitor of DNA

methylation, or trichostatin A, a histone deacetylase inhibitor, but neither induced GFP

expression (data not shown). Thus, the integrated lentiviral vector may have been lost

during the immortalization of the cells.

In order to determine if loss of Fps expression might have influenced the relative

expression or localization of N- and E-cadherin in these immortalized fps-null cells, we

rescued Fps expression using a murine stem cell virus (MSCV)-based retrovirus encoding

a myc-epitope tagged Fps fusion protein (Fps-myc). Immunoblotting analysis showed

that the infected cells expressed a distinct band that corresponded to the Fps-myc protein

(Figure 3.8D). However, subsequent immunofluorescence analysis of these rescued cells

revealed the same localization of N- and E-cadherin (data not shown), suggesting that

disrupted localization of these cadherins was not caused by loss of Fps expression.

3.5 Discussion

Fps expression in the mammary gland increases during pregnancy and then

reaches a maximum during lactation 107 (Truesdell and Greer, unpublished). Based upon

128

Figure 3.8 Confocal immunofluorescence microscopy analysis of the fps-null immortalized cells. A confluent monolayer of fps-null cells was stained with an antibody to E-cadherin (A), N-cadherin (B), and β-catenin (C) and detected with an Alexa-488 (green) or Alexa-633 (red) conjugated secondary antibody. Nuclei were stained with Hoechst 33258 (blue). In (D) lysates from fps-null cells infected with the MSCV-Fpsmyc retrovirus were immunoblotted with Fps/Fer antibody.

129

in situ analysis, this expression is primarily specific to epithelial cells (Truesdell and

Greer, unpublished). Recently it was demonstrated that the loss or inactivation of the Fps

tyrosine kinase resulted in earlier mammary tumor onset when compared to control mice

55, suggesting that Fps has a tumor suppressor function in the mammary gland. Therefore,

there is evidence that implicates Fps as a regulator of both normal mammary function and

tumorigenesis.

Many common cell lines of mammary origin that are used in biochemical studies

express either very little or undetectable amounts of the Fps protein (Truesdell and Greer,

unpublished). To study the function of Fps in the mammary gland our initial studies

involved the isolation of primary epithelial cells. However, primary cells have a finite

lifespan when grown in culture and typically enter senescence or die after a certain

number of generations 391, 410. One of the important regulators of these events is the p53

tumor suppressor. In vivo, p53 induces growth arrest or apoptosis of abnormal or

damaged cells, thereby preventing the development of cancerous cells. Due to this fact,

the p53 pathway is often inactivated in cancer cells through mutation or deletion 418. In

primary cell culture systems, the non-physiological growth conditions can similarly result

in the activation of the p53 pathway and inhibit growth or induce apoptosis 391, 410. Thus,

p53 is a logical, and likely a critical target in the generation of established cell lines from

primary cell cultures. In support of this hypothesis, the p53 knockout mouse has been

used to generate a number of different cell lines 419-421, suggesting that the loss of p53 is

important to extending the proliferative potential of cultured cells.

130

Using a lentiviral vector system that encodes a murine p53 shRNA, we provide

evidence that suppression of p53 expression in primary murine mammary epithelial cells

expands their growth potential in culture and leads to immortalization. A similar

lentiviral system that was employed in our study has previously been used to “knock-out”

p53 expression in senescent mouse embryo fibroblasts, resulting in cell cycle re-entry and

immortalization 415. This suggests that the use of shRNA is a convenient and efficient

strategy to suppress p53 expression.

Based on the expression of the GFP reporter, the initial percentage of cells that

could be infected in primary mammary gland cultures was low (less than 20 %). This was

surprising due to the documented ability of lentiviruses to infect non-dividing cells 422.

Using the same sample of virus, a very high percent of mouse embryonic fibroblasts

expressed GFP after infection, proving that the virus was produced at a relatively high

titer. In our experiments, the primary epithelial cells generated in the collagen gel were

isolated and plated on a fresh collagen gel as compact clumps. It appeared that densely

compacted cells were less susceptible to infection. If the cells had been less compact and

better dispersed there may have been a higher rate of infection. Despite the low level of

initial infection, within a few population doublings the majority of the remaining cells

were GFP positive. Although this is an indirect measure of the transcription of the p53

shRNA construct, and therefore the suppression of p53 expression, the fact that all cells

that survived several passages were positive for GFP indicates that p53 levels were

probably targeted as expected. When other primary cell preparations were maintained as

uninfected cells or infected with a control GFP-expressing lentivirus there was very little

131

epithelial cell survival beyond two or three passages. This further supports our belief that

cell survival was related to the lentiviral shRNA-mediated suppression of p53 expression.

The growth of primary cells in culture can lead to phenotypic and gene expression

changes. To assess whether the fps-null cells possessed epithelial characteristics, they

were examined for cell shape and protein biomarker expression. Occasionally small

groups of cells that appeared to have an elongated shape would grow within the epithelial

cells. These cells were never analyzed but their shape was indicative of a non-epithelial

phenotype. Accordingly, efforts were made to remove these cells from the remaining

cuboidal shaped epithelial cells. Confocal analysis with an E-cadherin antibody

confirmed that an important epithelial marker was still expressed and found in its normal

subcellular location. Similarly, β-catenin localized to cell-cell junctions. The cells were

negative for SMA, a marker for myoepithelial cells. These results suggest that the cells

have retained some of their epithelial features during their isolation and propagation.

With continued growth, an immortalized cell population eventually grew out of

the mixture of epithelial cells. These cells displayed a cuboidal shape which was

suggestive of an epithelial-like phenotype. When grown to 100 % confluence the

cuboidal shape was even more obvious and the cells displayed contact inhibition. Despite

the normal appearance of the cells, immunoblotting analysis revealed that in addition to

E-cadherin the cells also expressed N-cadherin. When epithelial cells undergo an

epithelial-to-mesenchymal transition (EMT) there is often a switch in cadherin expression

423. During an EMT, cells can eventually become transformed where their morphology

changes to a more spindal shape. The N-cadherin expression in the fps-null cells suggests

132

that, to some extent, such a transition may have occurred. The aberrant expression of a

second cadherin can competitively deplete the amount of p120-catenin that is able to

associate with the endogenous cadherin 261. The uncoupling of p120-catenin from E-

cadherin can result in its internalization and degradation 255.The examination of E-

cadherin by immunofluorescence revealed that the majority is located in cytoplasmic

vesicles. This observation coupled with the reduction in E-cadherin expression, when

compared to that in HC11 cells, implies that this process is likely taking place.

It was interesting that the HC11 cells expressed a significant amount of the

myoepithelial marker, SMA, while it was minimal in the fps-null cells. Conversely, the

expression of vimentin was abundant in the fps-null cells. Vimentin is the major

intermediate filament (IF) protein of mesenchymal cells 424, and its expression is a strong

indication that the cells have acquired fibroblast features.

If the p53 shRNA lentivirus functioned as predicted then p53 levels should be

reduced or absent in the transduced fps-null cells. HC11 cells are known to express an

inactive form of p53 186, and by immunoblotting, a single 53 kDa band was observed in

lysates from uninfected cells; whereas this band was not detected after transduction of

HC11 cells with the p53 shRNA lentivirus. Similarly, no band corresponding to p53 was

observed in the transduced fps-null cells. However, when the later passage fps-null cells

were examined by immunofluorescence, no GFP was detected. Expression of transgenes

can be down-regulated though epigenetic processes such as DNA methylation and

histone deacetylation 425. Therefore, the cells were treated with methylase and

deacetylase inhibitors in attempts to reactivate GFP expression. Neither drug was able to

133

do so. Currently we are unsure if the transgene is present in the fps-null cells. Although

the GFP and p53 shRNA sequences are within the same lentivirus integration cassette,

each is transcribed independently from different promoters. Despite the inactivation or

loss of GFP, it is possible that the p53 shRNA is still produced and functioning to

knockdown p53 expression. Indeed, immunoblotting analysis showed that these cells did

not express detectable levels of p53. The end result of this process was the generation of

a fps-null mammary cell line. And, with the use of the MSCV retroviral system we were

able to generate derivative cells that stably expressed a myc-epitope tagged Fps protein.

Both cell lines were tested for their ability to undergo terminal differentiation in the

presence of lactogenic hormones. However, based on the absence of casein production,

we conclude that neither cell line retained the ability to differentiate into functional

casein-producing epithelial cells in culture (data not shown). In contrast, HC11 cells did

form characteristic dome-like structures and produced casein proteins in comparable

cultures (data not shown).

Because the cells have probably undergone an EMT process, it was possible they

possessed properties that were acquired during this phenotypic change. Epithelial cells

that have acquired a fibroblast-like phenotype can be more motile and display anchorage

independent growth 426. To test the migration of the cells the Boyden chamber method

was used. No migration was observed when the attractant was 10 % FBS supplemented

with 20 ηg/ml EGF. Minimal migration was induced when the cells were stimulated with

1 ηg/ml HGF, but it was a very low fraction of the plated cells (< 0.01 %) (data not

shown). Therefore, despite the presence of EMT markers including N-cadherin and

134

vimentin, the cells were not migratory. Similarly, no colony formation occurred when the

cells were grown in soft agar (data not shown). Thus, despite the apparent mesenchymal

transition of these cells, they were not invasive or transformed.

The lack of normal mammary cell characteristics observed in the cells suggests

that they will have limited use for studying the role of Fps in the context of normal

mammary gland functions associated with lactation. However, Fps also has a role in

tumor development which may be unlinked to functions in lactation. The continued

growth or introduction of appropriate oncogenes may be sufficient to promote the

transformation of the cells such that they become a suitable model system to investigate a

potential intrinsic tumor suppressor role of Fps.

135

Chapter 4

General Discussion

The main objective of my thesis was to determine the function(s) of the Fps

tyrosine kinase in the mouse mammary gland. There were two lines of evidence

indicating that Fps had a function in the mammary gland. First, fps transcripts were

shown to be differentially regulated during mammary development. Specifically,

expression was up-regulated during later stages of pregnancy and then reached a maximal

level during lactation 107. Second, in a MMTV/PymT breast tumor mouse model the

targeted deletion or inactivation of Fps correlated with a significantly earlier tumor onset

than in wild-type counterparts 55. Together, these results led us to investigate the

biochemical and physiological roles of Fps in mammary gland homeostasis and function.

In this study, we have chosen to investigate the role of Fps in the mammary gland with

the use of the fps-null mouse line which lacks Fps protein expression due to a deletion of

exons 15-18 26.

Chodosh et al., provided some interesting information on the changing levels of

Fps expression in the mammary gland during the reproductive cycle but the data shown

were far from conclusive 107. Because of the faint northern blot signal and the fact that

transcript levels do not always directly correlate with protein expression level a more in

depth analysis of Fps expression was warranted. To provide more definite information on

Fps expression we examined Fps protein levels in wild-type animals at different

developmental stages. In agreement with the pattern of fps mRNA expression, we have

shown that the Fps protein was expressed at all stages of mammary development. It then

136

became up-regulated at later stages of pregnancy and most highly expressed during

lactation. These results and those from the Leder lab are the only reports describing the

expression of Fps in the mammary gland 107.

The phosphorylation of Fps and presumptive kinase activation observed in the

mammary gland during lactation was quite surprising and intriguing. This level of Fps

phosphorylation was compared to that in; 1) other tissues, 2) other stages of mammary

development, or 3) two cell lines that over-express the wild-type Fps or an activated

myristoylated Fps variant. While there was no apparent Fps activity in any of the other

tissues tested, Fps phosphorylation was detected in the cell lines, with more activity

associated with activated Fps. However, it was still significantly below the level observed

in the lactating mammary gland. This suggests that Fps becomes robustly phosphorylated

during lactation and its activation exceeds the level that can lead to transformation in

some types of cells 51. The significance of this is not known but it is likely to be

important to its normal function in the mammary gland during lactation.

The activity of Fps is believed to be tightly repressed under normal physiological

conditions 349. However, it can become abundantly autophosphorylated when over-

expressed in cultured cells, which would indicate that the level of expression may be a

factor in Fps activation. This may be true in certain circumstances but the relatively

higher level of activation in the mammary gland compared to Fps-expressing cultured

cells suggests that other unknown factors are involved.

The kinase activity of Fps is thought to be dependent on its ability to oligomerize

in a homotypic fashion 85. Oligomerization is believed to occur via the second predicted

137

coiled-coil (CC) motif which, in turn, leads to in trans autophosphorylation 86, 351.

Conversely, the first CC motif within the putative F-BAR domain may have a role in

negatively regulating the activity of Fps, which is supported by the observed increase in

activity when a point mutation was introduced into it 351. These authors have suggested

that the first CC motif may interact with the second CC motif and, in doing so, block the

ability of the second CC motif to mediate oligomerization 87. Assuming this model is

correct, then upstream effectors of Fps kinase activation must somehow cause the release

of the negative effects of the first CC motif. The recruitment of Fps to sites of activation,

such as cytokine or growth factor receptors or membrane surfaces, may be sufficient to

bring multiple molecules of Fps into close proximity and lead to activation. The ability of

the Fps F-BAR domain to preferentially interact with specific forms of membrane lipid

molecules could also be involved in its activation. Different phosphoinositides can be

localized to specific regions of a membrane 427 and could theoretically regulate the extent

that Fps molecules come into contact with one another. If inactive Fps exists as a

monomer in a closed configuration, then the accumulation of monomers in close

proximity could promote the interaction of F-BAR domains in trans, causing a

conformational shift within Fps and relieving the inhibitory effects of the first CC motif.

A similar model has been proposed to explain how the binding of the Fps SH2 domain to

pY-containing substrates generates a shift in the relative position of the kinase domain,

enabling it to bind and phosphorylate the substrate (Tony Pawson, unpublished).

Regardless of the precise mechanism that leads to Fps activation, it is almost certain to

involve the N-terminal domain. This is supported by the observations that N-terminal

138

myristoylation or fusion to the viral Gag protein results in deregulated kinase activity 12,

13, 51.

Fps is known to have a relatively restricted expression pattern. It has been found

in a variety of cell types within the hematopoietic system including platelets,

erythrocytes, mast cells, macrophage and granulocytic cells (reviewed in 5). Early studies

suggested that Fps expression was restricted to the myeloid lineage hematopoietic system

but it is now known to be expressed in endothelial, neuronal and various types of

epithelial cells 52, 54, 99, 347, 428. The presence of fps transcripts has been reported in a

transformed murine mammary epithelial cell line 52, but immunoblot analysis failed to

detect any Fps protein (Waheed Sangrar, unpublished).

To gain some insight into the types of cells that expressed Fps we performed

immunofluorescence microscopy analysis on lactating mammary gland tissue using cell-

type specific antibodies in conjunction with anti-Fps antibodies. By comparing the cells

that expressed an epithelial-specific (Mucin-1) or myoepithelial-specific (smooth muscle

actin) protein to Fps-positive cells, we have shown that Fps is prominently expressed in

mammary epithelial cells. That Fps was detected in the epithelial compartment of the

mammary gland was not an unrealistic expectation. However, the fact that the level of

Fps expression increased throughout pregnancy and then increased dramatically as the

mammary gland transitioned to the lactation phase was surprising. This suggests that the

changes in expression levels are actively regulated and are likely a reflection of the

differentiation and functional state of the gland. In the hematopoietic system, Fps is also

preferentially more highly expressed in differentiated myeloid cells, eg. monocytes and

139

granulocytes, than in less differentiated cells 429. The transcription factors, Sp1 and PU.1,

have been identified as transactivators of fps expression in these cell types 30, 31. Whereas

PU.1 is specific to the hematopoietic system, Sp1 is ubiquitously expressed, participates

in the transcriptional regulation of hundreds of genes, and may participate in activating

fps expression in the cells where fps is normally expressed 34. Sp1 is known to regulate

the transcription of some genes, such as connexin 26, in the mammary gland that are up-

regulated during pregnancy and lactation 430. Connexin 26 is also regulated by the AP-2

transcription factor 431. Interestingly, the fps promoter contains at least three predicted

AP-2α binding sites (Zirngibl, unpublished). It is not known if AP-2α actually can bind

these sites or activate fps expression but its presence in the mammary gland at least

makes it a possibility 432. However, although AP-2α is expressed throughout the

developmental cycle of the mammary gland 432, its temporal pattern of expression seems

to inversely correlate with that of fps. It is possible that AP-2α functions to negatively

regulate fps expression, and its reduction during lactation enables the increase of fps

expression.

There are also three predicted nuclear factor (NF)-1 binding sites in the fps

promoter (Zirngibl, unpublished). Again, nothing is definitively known about this

transcription factor influencing fps expression but some members of the NF-1 family are

expressed in the mammary gland and NF-1-C2 can activate gene-specific expression at

mid-pregnancy 433, which is also the approximate time when fps expression begins to be

up-regulated. None of the transcription factors that are known or predicted to bind the fps

140

promoter have been shown to be more active during lactation, so the factors that regulate

the increase in fps expression during this stage remain a mystery.

When comparing the types of cells that express abundant levels of Fps, it is

apparent that each possesses prominent secretory capabilities 52. Interestingly, milk-

producing mammary epithelial cells, where Fps is most highly expressed, are probably

one of the most highly active secretory cells. The expression of fluorescently tagged Fps

proteins has been used to visualize Fps in a perinuclear region that was consistent with

the Golgi apparatus 63. Molecular Markers for the trans Golgi network (TGN38 and γ-

adaptin), the ER-Golgi intermediate compartment (Rab1), the exocytic pathway (Rab3A),

and the early and late endocytic pathways (Rab5B and Rab7, respectively) were found to

co-localize with Fps 52, 63. This suggests that Fps may be involved in regulating multiple

components of the secretory and endocytic pathways. Additional evidence for this theory

has recently been generated in our lab using time-lapse fluorescence microscopy where a

transiently expressed Fps-dsRed fusion protein was localized to multiple vesicular

structures in different compartments of the cell (Jonathan Zipursky, unpublished). In the

mammary gland, Fps displayed a similar punctate pattern seen in other cell types, though

it was not concentrated perinuclearly as previously described. This may be related to the

tremendous expansion of cytoplasmic secretory organelle content of mammary epithelial

cells that accompanies the onset of lactation 175, 176.

The mechanism by which Fps associates with secretory structures is not known.

However, it is tempting to speculate that the putative N-terminal F-BAR domain could be

involved. The F-BAR domain is believed to be responsible for membrane binding and the

141

generation of membrane curvature during such biological processes as filopodia

formation and endocytosis 80. This is true for several of the F-BAR domain proteins but

may not be the case with respect to the Fps or Fer kinases, as over-expression of the

putative Fer F-BAR domain failed to induce substantial tubulation 67. However, putative

Fps and Fer F-BAR domains are capable of binding lipids and selectively interacting with

different lipid species (Andrew Craig and Michelle Scott, unpublished) 67. Specifically,

Fps preferentially binds to PI3P, PI4P, PI5P, and PI(4,5)P2. These compounds have been

identified on distinctive intracellular organelles due to their finely controlled synthesis

and metabolism by PI-kinases and PI-phosphatases 434. For instance, PI4P has been

shown to be more prominent in the Golgi than at the plasma membrane and, even within

the Golgi, there is twice as much in early Golgi compartments compared with that of the

trans Golgi network 435. PI(4,5)P2, although not more prominent, has been implicated in

Golgi intracisternal transport and vesicle formation from the trans Golgi network 434, 436.

The ability of the different PIPs to interact with PI-binding domains enables them to

serve as docking sites for specific proteins. The non-uniform distribution of the different

PIP species provides a mechanism for the targeted recruitment and activation of protein

effectors. It is possible that Fps gets recruited by localized concentrations of specific PIPs

to different secretory or endocytic compartments where it can undergo homotypic

oligomerization and autophosphorylation, followed by interaction with and

phosphorylation of downstream effectors.

The differences in the weights of the pups reared by fps-null and wild-type

females is probably the most compelling evidence to suggest that Fps has a biological

142

function in the mammary gland. As it is unlikely that Fps contributes to mammary

development or differentiation it is possible that Fps regulates a lactation function of the

mammary gland which is compromised in the fps-null females. When trying to ascribe a

function to Fps in milk production it is important to recognize that there are two primary

routes of secretion that account for most of the milk components 203. The exocytotic

pathway is the primary route of secretion for such compounds as proteins, water, lactose,

oligosaccharides, phosphate, citrate and calcium; whereas the milk fat, primarily

triacylglycerides, is secreted by the lipid pathway.

Once the molecules that are to be secreted via the exocytotic pathway enter, or are

produced in, the Golgi lumen, they transition through the Golgi network where they are

gradually modified, sorted and concentrated before being released in a secretory vesicle.

This process involves a series of vesicle formations and fusions between adjacent Golgi

cisternae and the eventual release of the secretory vesicle toward the plasma membrane.

Vesicle formation involves both the curvature and severing of the donor membrane 437,

438. The presence of a number of BAR and F-BAR domain-containing proteins in the

Golgi network has been demonstrated. In conjunction with other lipid-binding and

membrane-deforming proteins, they participate in the vesiculation of sequestered cargo in

the Golgi network 73. Additionally, many of the proteins, or variants of these proteins,

that are involved in the severing or pinching of the vesicle from the plasma membrane

during endocytosis localize to the Golgi and probably provide a similar function.

Tyrosine phosphorylation of a number of these proteins is known to occur and may

influence their activity and/or ability to bind interacting partners. Co-expression studies

143

have shown that Fps can phosphorylate members of the PACSIN family (Yan Gao,

unpublished), a group of F-BAR domain-containing proteins that have been implicated in

vesicular trafficking and regulation of dynamin-mediated endocytosis 70. This shows that

some of the molecules that are involved in vesicular formation are potential in vivo

substrates for tyrosine kinases like Fps. Whether Fps contributes to the exocytotic

pathway is not known, but it is an interesting possibility worthy of future study.

The process by which milk lipid globules are secreted by epithelial cells is

believed to be a unique pathway for lipid secretion 439. Milk fat globules originate as

small droplets of triacylglerides that are enzymatically produced on or in the membrane

of the smooth endoplasmic reticulum (SER). The lipids are released from the SER

through an unknown mechanism into the cytoplasm as microlipid droplets that are coated

with protein and polar lipid from the outer half of the SER membrane 175. A range of lipid

droplet sizes is found in the cytoplasm due to their ability to fuse with one another.

Independent of their size, the droplets transit to the apical membrane, interact with and

become enveloped by the plasma membrane before being released into the milk. Because

so little information is known about the molecular mechanisms involved in milk fat

globule formation and secretion it is difficult to propose a specific biochemical role for

Fps in this process.

A feature that is shared by both the exocytotic and lipid pathways is the

dependence on the cytoskeleton for transporting vesicles and lipid droplets to the

membrane for secretion. Here again, not much is known about the details of each process

in mammary epithelial cells but both the microtubule and microfilament systems are

144

likely to be involved 206, 208. Secretion via the exocytotic pathway can be blocked by

treatment with cytochalasin B, a microfilament-altering drug 211, or nocodazole and

colchicine, microtubule depolymerizing agents 209, 210. Indirect evidence suggests that the

lipid pathway also depends on the microtubule system. An analysis of the proteins that

were associated with cytoplasmic lipid droplets identified several components of the

microtubule motors and cytoskeletal-interacting proteins, including dynein intermediate

chain, motor protein, gelsolin and gephyrin 219. These proteins are associated with the

transport of cargo towards the microtubule minus-end. Because of this fact, it is likely

that the lipid droplets are transported along microtubules which have their minus-ends

oriented toward the apical surface. The significance of this is not completely understood,

in part because much of the framework for considering how microtubule organization is

regulated is based on results from cells where microtubule minus-ends are organized

around the centrosome and the plus-ends are directed towards the cell periphery 440.

However, it appears that acentrosomal microtubule minus-end orientation toward the

apical surface is common in epithelial cells 222, 441, 442. Moreover, apical cargo transport in

Madin-Darby canine kidney cells has been shown to involve both the dynein and kinesin

motors 208, suggesting that a combination of minus-end and plus-end directed

microtubule-associated motors are involved in secretion. In general, microtubules in

epithelial cells are considered more stable than those in fibroblast-like cells 443, with a

slower rate of growth and shortening, and a lower overall dynamicity 441.

There is evidence to suggest that Fps, and possibly Fer, have a role in regulating

the microtubule system, at least in some cell types. Little is known about the role of Fer,

145

but in polarized endothelial cells it has been found associated with microtubules growing

toward cell-cell contacts 444. Conversely, more is known about the role of Fps.

Specifically, Fps can interact with and phosphorylate α-tubulin 82, 83, 347, which contains a

tyrosine residue within the sequence (YEEV) that conforms to the Fps kinase consensus

sequence. The binding of Fps to soluble tubulin requires the FCH domain 82, 83. Similarly,

Fps can bind stable microtubules but this interaction is dependent on kinase activity and

an intact SH2 domain 82, 83. These studies have also revealed that Fps may be involved in

microtubule nucleation and bundling, as suggested by the co-localization of Fps with γ-

tubulin and the Fps-mediated tubulin polymerization in vitro 82, 83. Despite the expected

low level of fps expression in fibroblast cells, mouse embryonic fibroblasts from fps-

deficient mice displayed reduced stable microtubules, as well as decreased bundling and

nucleation 82, 83. Together, these results implicate Fps as a regulator of the tubulin

cytoskeleton.

The fact that an intact microtubule system is important for the secretion of milk

components and Fps is highly expressed in mammary epithelial cells during lactation

evokes the possibility that Fps may have a role in microtubule regulation. The ability of

Fps to phosphorylate tubulin and induce microtubule bundling suggests that in mammary

epithelial cells it could contribute to the establishment and/or maintenance of the

microtubule system. Thus, fps-null mammary epithelial cells may contain an altered or

reduced system of microtubules that leads to a decreased secretion rate for components of

both the exocytotic and lipid pathways. Despite this possibility, a functional microtubule

system must still exist in these cells because milk is produced for the pups to grow and

146

survive. But microtubules are dynamically unstable structures that are proposed to never

reach a steady state 440; instead existing in alternating states of polymerization and

depolymerization for various times depending on the conditions. If the mechanisms

controlling the microtubule length or rate of generation/breakdown become deregulated,

then the biological processes that depend on proper function of these structures could be

negatively affected. Thus, in the absence of Fps it is possible that microtubules form at a

slower rate or get broken down more quickly. This would result in a slower or delayed

movement of secretory vesicles toward the apical plasma membrane and a reduction in

the total amount of milk produced over a given time period. Thus, even in the absence of

obvious differences in milk protein or fat content, this type of defect could contribute to

the weight differences observed in pups reared by fps-null and wild-type mothers.

The dramatic increase in Fps kinase activity during lactation suggested that it

likely interacted with and phosphorylated one or more proteins in the mammary gland

during this stage of development. To identify interacting partners and potential substrates,

protein candidates were chosen whose altered function could promote tumorigenesis and

result in abnormal mammary function. Initial experiments focused on E-cadherin and

other components of the adherens junction (AJ) complex. Because the Fer kinase is

known to be a component of both N- and E-cadherin based AJs 321, 334 it led to the

speculation that Fps could also regulate the AJ in the mammary gland. The role of Fer in

AJs has been extensively studied in the context of N-cadherin-based AJs but it is not

known whether Fer is an AJ component for all classical cadherins or in all cell types 321.

Additionally, Fps and Fer are known to have some functional redundancy in other cell

147

systems 49. Thus, it was possible that Fps could replace or compliment the function of Fer

in the mammary gland. Analysis of tissues from different developmental stages showed

that Fps was a component of the E-cadherin-based AJ complex. This interaction was

observed in different stages of pregnancy but was most obvious during lactation. Not

surprising, Fps was identified in association with β-catenin and p120-catenin. Although

Fer was also present in the AJ complex, Fps was more predominant; suggesting that Fps

may have a more prominent role than Fer in regulating mammary epithelial cell-cell

adhesion. It is interesting that both Fps and Fer were found in the AJ, and may be

indicative of unique function(s) for each kinase.

The decision to focus on the components of the AJ as possible Fps substrates was

partially due to the fact that Fer has a well established role in regulating the

phosphorylation of multiple AJ proteins, and we speculated that Fps could compliment or

replace the function of Fer in mammary epithelial cells. The presence of Fer in the AJ is

important to the overall stability of the cadherin complex and is dependent on its

interaction with p120-catenin 321, 334. When bound to p120-catenin Fer is then able to

directly phosphorylate β-catenin on Y142 causing its interaction with α-catenin to be

disrupted 320. Fer has also been identified as an important regulator of the

PTP1B/cadherin interaction 321. The phosphorylation of PTP1B on Y152, which is

required for its ability to interact with cadherins, is dependent on Fer activity 321. In turn,

PTP1B acts on Y654 of β-catenin to maintain its dephosphorylation and preserve the β-

catenin/cadherin interaction. In cells that lack Fer kinase activity, PTP1B and β-catenin

148

become uncoupled from N-cadherin, with N-cadherin and β-catenin being lost from cell-

cell contacts 321.

It is doubtful that Fps assumes the role of Fer in the E-cadherin-based AJ. In

support of this is the fact that the interaction between E-cadherin and β-catenin was

normal in fps-null tissue, and they localized to cell-cell junctions. Also, there was no

tyrosine phosphorylation of β-catenin detected in the fps-null tissues. In fact, none of the

AJ core components displayed any tyrosine phosphorylation regardless of the presence or

absence of Fps. This is probably the strongest evidence that eliminates E-cadherin, β-

catenin, p120-catenin and PTP1B as substrates of Fps, at least in the lactating mammary

gland. A puzzling aspect of the presence of Fps in the AJ is that it is not tyrosine

phosphorylated, or it is below the level of detection. This result is difficult to explain but

may indicate that Fps possesses non-kinase activities. It is also possible that in mature

AJs, where very little turnover is occurring at a given time, only a small fraction of the

total Fps is transiently activated at a given time. This would likely make detecting Fps

phosphorylation very difficult.

The ability of p120-catenin to act as a docking protein for Fer and the Src-family

kinases, Yes and Fyn 320, suggests that it could recruit other cytoplasmic kinases to the

AJ. However, when the interaction of Fps with the individual components was examined,

it did not seem to be more strongly associated with p120-catenin. Surprisingly, a similar

pattern was observed for Fer. Thus, we cannot exclude the possibility that Fps directly

interacts with p120-catenin but it remains to be conclusively determined with which

protein(s) Fps directly binds in the AJ. Nevertheless, we have provided evidence that

149

implicates Fps in regulating the interaction between p120-catenin and E-cadherin/β-

catenin. Specifically, less E-cadherin and β-catenin are associated with p120-catenin in

the absence of Fps. These data have led to our proposed model where the presence of Fps

in the E-cadherin-based AJ serves to stabilize the interaction of p120-catenin with E-

cadherin/β-catenin (Figure 4.1). One of the important functions of p120-catenin in the AJ

is the stabilization of the cadherin complex at the cell surface 254, 255. When p120-catenin

is reduced or deleted the AJ complex undergoes targeted internalization, where it may get

recycled or degraded 254, 255. Typically, cadherin staining is lost or mislocalized in p120-

catenin deficient cells 254, 255. In fps-null cells, E-cadherin and β-catenin are still found at

cell-cell junctions somewhat similar to wild-type cells but the staining pattern seems to

be more disorganized and less focused.

The affinity of p120-catenin for cadherin has been shown to be relatively weak

252. Our results that show low levels of p120-catenin in IPs of either E-cadherin or β-

catenin support this idea. This characteristic of p120-catenin is likely important to its

ability to rapidly alternate between cadherin-bound and -unbound states. It could also be

required as part of a cells response to different stimuli where stable AJs are disassembled

and levels of free p120-catenin increase. Any alteration to the make-up of the AJ could

theoretically shift the balance toward increased levels of unbound p120-catenin. The

ability of p120-catenin to interact with multiple binding partners 261 could make it

susceptible to aberrant protein interactions where, in the absence of Fps, p120-catenin is

competitively removed from the AJ.

150

Figure 4.1 Proposed model for the role of Fps in E-cadherin-based AJ in the mouse mammary gland during lactation. In the presence of Fps, the AJ components form stable complexes at distinct regions of cell-cell contacts between adjacent cells. In the absence of Fps, p120-catenin becomes uncoupled from E-cadherin and β-catenin and destabilizes the complex.

151

If it can be established that Fps has a more general role in the regulation of AJ function,

then this may have important implications in the development of cancer. Our lab has

previously shown that mammary tumor development occurs earlier in the absence of Fps

55. Also, Fps mutations have been identified in several colon cancer samples 53 that were

subsequently shown to cause kinase inactivation 54, 55. These data suggest that Fps may

act as a tumor suppressor in certain cell types. When AJs from adult unmated mice were

examined for the presence of Fps, the results were inconsistent but with immunoblot

analysis we were able to detect a faint 92 kDa band that corresponded to Fps (data not

shown). If this result can be substantiated then it would provide evidence that Fps

interacts with tumor suppressor proteins with well established functions.

Each of the core components of the AJ has an important role in maintaining

proper cell-cell adhesion (reviewed in 445. AJ components can be disrupted through a

number of different mechanisms, including post-translational modification, mutation,

deletion, reduced or lost expression and mislocalization 280-285. If the loss of Fps is

responsible for disrupting the interaction between p120-catenin and E-cadherin then this

may be a possible explanation for the earlier occurrence of mammary tumors in the mice

with fps-null and fps-inactivating mutations. It has been suggested that because of the

role of p120-catenin in stabilizing E-cadherin at the membrane, the loss of p120-catenin

function may have a more prominent role in disrupting the AJ than previously realized;

especially in the cancer cases which display an apparent lack of alteration in E-cadherin

expression 261.

152

In summary, the results presented in this thesis suggest that Fps has an important

function in the mammary gland in which, in its absence, an aspect of lactation is

compromised. The increase in Fps expression and activation during lactation, along with

the reduced weight of pups reared by fps-null females, implicate Fps as a regulator of

milk production and/or secretion. During lactation, there is a strong association between

Fps and components of the E-cadherin-based AJ. It appears that the normal stoichiometry

of the AJ is altered in fps-null cells. The effect that this difference has on the phenotype

of the mammary gland is not yet fully understood. But the role of the E-cadherin-based

AJ in normal mammary function should not be underestimated, as it is critical to

epithelial cell shape, polarity, differentiation and homeostasis 360. The results presented

here should provide a foundation for future studies that will identify the specific

biological and molecular function(s) of Fps in the mouse mammary gland.

153

References

1. Franchini G, Even J, Sherr CJ, Wong-Staal F: onc sequences (v-fes) of Snyder-Theilen feline sarcoma virus are derived from noncontiguous regions of a cat cellular gene (c-fes), Nature 1981, 290:154-157 2. Shibuya M, Hanafusa H: Nucleotide sequence of Fujinami sarcoma virus: evolutionary relationship of its transforming gene with transforming genes of other sarcoma viruses, Cell 1982, 30:787-795 3. Shibuya M, Hanafusa T, Hanafusa H, Stephenson JR: Homology exists among the transforming sequences of avian and feline sarcoma viruses, Proc Natl Acad Sci U S A 1980, 77:6536-6540 4. Groffen J, Heisterkamp N, Shibuya M, Hanafusa H, Stephenson JR: Transforming genes of avian (v-fps) and mammalian (v-fes) retroviruses correspond to a common cellular locus, Virology 1983, 125:480-486 5. Greer P: Closing in on the biological functions of Fps/Fes and Fer, Nat Rev Mol Cell Biol 2002, 3:278-289 6. Lee WH, Phares W, Duesberg PH: Structural relationship between the chicken DNA locus, proto-fps, and the transforming gene of Fujinami sarcoma virus, delta gag-fps, Virology 1983, 129:79-93 7. Roebroek AJ, Schalken JA, Onnekink C, Bloemers HP, Van de Ven WJ: Structure of the feline c-fes/fps proto-oncogene: genesis of a retroviral oncogene, J Virol 1987, 61:2009-2016 8. Hammond CI, Vogt PK, Bishop JM: Molecular cloning of the PRCII sarcoma viral genome and the chicken proto-oncogene c-fps, Virology 1985, 143:300-308 9. Bowzard JB, Visalli RJ, Wilson CB, Loomis JS, Callahan EM, Courtney RJ, Wills JW: Membrane targeting properties of a herpesvirus tegument protein-retrovirus Gag chimera, J Virol 2000, 74:8692-8699 10. Pawson T, Guyden J, Kung TH, Radke K, Gilmore T, Martin GS: A strain of Fujinami sarcoma virus which is temperature-sensitive in protein phosphorylation and cellular transformation, Cell 1980, 22:767-775 11. Sadowski I, Pawson T, Lagarde A: v-fps protein-tyrosine kinase coordinately enhances the malignancy and growth factor responsiveness of pre-neoplastic lung fibroblasts, Oncogene 1988, 2:241-247 12. Meckling-Gill KA, Yee SP, Schrader JW, Pawson T: A retrovirus encoding the v-fps protein-tyrosine kinase induces factor-independent growth and tumorigenicity in FDC-P1 cells, Biochim Biophys Acta 1992, 1137:65-72 13. Carmier JF, Samarut J: Chicken myeloid stem cells infected by retroviruses carrying the v-fps oncogene do not require exogenous growth factors to differentiate in vitro, Cell 1986, 44:159-165 14. Hjermstad SJ, Briggs SD, Smithgall TE: Phosphorylation of the ras GTPase-activating protein (GAP) by the p93c-fes protein-tyrosine kinase in vitro and formation of GAP-fes complexes via an SH2 domain-dependent mechanism, Biochemistry 1993, 32:10519-10525 15. Ellis C, Moran M, McCormick F, Pawson T: Phosphorylation of GAP and GAP-associated proteins by transforming and mitogenic tyrosine kinases, Nature 1990, 343:377-381

154

16. Moran MF, Koch CA, Anderson D, Ellis C, England L, Martin GS, Pawson T: Src homology region 2 domains direct protein-protein interactions in signal transduction, Proc Natl Acad Sci U S A 1990, 87:8622-8626 17. Fukui Y, Saltiel AR, Hanafusa H: Phosphatidylinositol-3 kinase is activated in v-src, v-yes, and v-fps transformed chicken embryo fibroblasts, Oncogene 1991, 6:407-411 18. McGlade J, Cheng A, Pelicci G, Pelicci PG, Pawson T: Shc proteins are phosphorylated and regulated by the v-Src and v-Fps protein-tyrosine kinases, Proc Natl Acad Sci U S A 1992, 89:8869-8873 19. Kurata WE, Lau AF: p130gag-fps disrupts gap junctional communication and induces phosphorylation of connexin43 in a manner similar to that of pp60v-src, Oncogene 1994, 9:329-335 20. Maru Y, Peters KL, Afar DE, Shibuya M, Witte ON, Smithgall TE: Tyrosine phosphorylation of BCR by FPS/FES protein-tyrosine kinases induces association of BCR with GRB-2/SOS, Mol Cell Biol 1995, 15:835-842 21. Garcia R, Yu CL, Hudnall A, Catlett R, Nelson KL, Smithgall T, Fujita DJ, Ethier SP, Jove R: Constitutive activation of Stat3 in fibroblasts transformed by diverse oncoproteins and in breast carcinoma cells, Cell Growth Differ 1997, 8:1267-1276 22. Nelson KL, Rogers JA, Bowman TL, Jove R, Smithgall TE: Activation of STAT3 by the c-Fes protein-tyrosine kinase, J Biol Chem 1998, 273:7072-7077 23. Yee SP, Mock D, Greer P, Maltby V, Rossant J, Bernstein A, Pawson T: Lymphoid and mesenchymal tumors in transgenic mice expressing the v-fps protein-tyrosine kinase, Mol Cell Biol 1989, 9:5491-5499 24. Huang CC, Hammond C, Bishop JM: Nucleotide sequence and topography of chicken c-fps. Genesis of a retroviral oncogene encoding a tyrosine-specific protein kinase, J Mol Biol 1985, 181:175-186 25. Roebroek AJ, Schalken JA, Verbeek JS, Van den Ouweland AM, Onnekink C, Bloemers HP, Van de Ven WJ: The structure of the human c-fes/fps proto-oncogene, Embo J 1985, 4:2897-2903 26. Zirngibl RA, Senis Y, Greer PA: Enhanced endotoxin sensitivity in fps/fes-null mice with minimal defects in hematopoietic homeostasis, Mol Cell Biol 2002, 22:2472-2486 27. Wilks AF, Kurban RR: Isolation and structural analysis of murine c-fes cDNA clones, Oncogene 1988, 3:289-294 28. Greer P, Maltby V, Rossant J, Bernstein A, Pawson T: Myeloid expression of the human c-fps/fes proto-oncogene in transgenic mice, Mol Cell Biol 1990, 10:2521-2527 29. Heydemann A, Warming S, Clendenin C, Sigrist K, Hjorth JP, Simon MC: A minimal c-fes cassette directs myeloid-specific expression in transgenic mice, Blood 2000, 96:3040-3048 30. Heydemann A, Juang G, Hennessy K, Parmacek MS, Simon MC: The myeloid-cell-specific c-fes promoter is regulated by Sp1, PU.1, and a novel transcription factor, Mol Cell Biol 1996, 16:1676-1686 31. Ray-Gallet D, Mao C, Tavitian A, Moreau-Gachelin F: DNA binding specificities of Spi-1/PU.1 and Spi-B transcription factors and identification of a Spi-1/Spi-B binding site in the c-fes/c-fps promoter, Oncogene 1995, 11:303-313 32. Gupta P, Gurudutta GU, Verma YK, Kishore V, Gulati S, Sharma RK, Chandra R, Saluja D: PU.1: An ETS family transcription factor that regulates leukemogenesis besides normal hematopoiesis, Stem Cells Dev 2006, 15:609-617

155

33. McKercher SR, Henkel GW, Maki RA: The transcription factor PU.1 does not regulate lineage commitment but has lineage-specific effects, J Leukoc Biol 1999, 66:727-732 34. Lomberk G, Urrutia R: The family feud: turning off Sp1 by Sp1-like KLF proteins, Biochem J 2005, 392:1-11 35. Feldman RA, Tam JP, Hanafusa H: Antipeptide antiserum identifies a widely distributed cellular tyrosine kinase related to but distinct from the c-fps/fes-encoded protein, Mol Cell Biol 1986, 6:1065-1073 36. MacDonald I, Levy J, Pawson T: Expression of the mammalian c-fes protein in hematopoietic cells and identification of a distinct fes-related protein, Mol Cell Biol 1985, 5:2543-2551 37. Pawson T, Letwin K, Lee T, Hao QL, Heisterkamp N, Groffen J: The FER gene is evolutionarily conserved and encodes a widely expressed member of the FPS/FES protein-tyrosine kinase family, Mol Cell Biol 1989, 9:5722-5725 38. Hao QL, Heisterkamp N, Groffen J: Isolation and sequence analysis of a novel human tyrosine kinase gene, Mol Cell Biol 1989, 9:1587-1593 39. Fischman K, Edman JC, Shackleford GM, Turner JA, Rutter WJ, Nir U: A murine fer testis-specific transcript (ferT) encodes a truncated Fer protein, Mol Cell Biol 1990, 10:146-153 40. Plowman GD, Sudarsanam S, Bingham J, Whyte D, Hunter T: The protein kinases of Caenorhabditis elegans: a model for signal transduction in multicellular organisms, Proc Natl Acad Sci U S A 1999, 96:13603-13610 41. Cetkovic H, Muller IM, Muller WE, Gamulin V: Characterization and phylogenetic analysis of a cDNA encoding the Fes/FER related, non-receptor protein-tyrosine kinase in the marine sponge sycon raphanus, Gene 1998, 216:77-84 42. Katzen AL, Montarras D, Jackson J, Paulson RF, Kornberg T, Bishop JM: A gene related to the proto-oncogene fps/fes is expressed at diverse times during the life cycle of Drosophila melanogaster, Mol Cell Biol 1991, 11:226-239 43. Navon A, Schwarz Y, Hazan B, Kassir Y, Nir U: Meiosis-dependent tyrosine phosphorylation of a yeast protein related to the mouse p51ferT, Mol Gen Genet 1994, 244:160-167 44. Ludolf F, Bahia D, Andrade LF, Cousin A, Capron M, Dissous C, Pierce RJ, Oliveira G: Molecular analysis of SmFes, a tyrosine kinase of Schistosoma mansoni orthologous to the members of the Fes/Fps/Fer family, Biochem Biophys Res Commun 2007, 360:163-172 45. Feldman RA, Gabrilove JL, Tam JP, Moore MA, Hanafusa H: Specific expression of the human cellular fps/fes-encoded protein NCP92 in normal and leukemic myeloid cells, Proc Natl Acad Sci U S A 1985, 82:2379-2383 46. Keller P, Payne JL, Tremml G, Greer PA, Gaboli M, Pandolfi PP, Bessler M: FES-Cre targets phosphatidylinositol glycan class A (PIGA) inactivation to hematopoietic stem cells in the bone marrow, J Exp Med 2001, 194:581-589 47. Craig AW, Greer PA: Fer kinase is required for sustained p38 kinase activation and maximal chemotaxis of activated mast cells, Mol Cell Biol 2002, 22:6363-6374 48. Sangrar W, Gao Y, Bates B, Zirngibl R, Greer PA: Activated Fps/Fes tyrosine kinase regulates erythroid differentiation and survival, Exp Hematol 2004, 32:935-945 49. Senis YA, Craig AW, Greer PA: Fps/Fes and Fer protein-tyrosinekinases play redundant roles in regulating hematopoiesis, Exp Hematol 2003, 31:673-681

156

50. Care A, Mattia G, Montesoro E, Parolini I, Russo G, Colombo MP, Peschle C: c-fes expression in ontogenetic development and hematopoietic differentiation, Oncogene 1994, 9:739-747 51. Greer P, Haigh J, Mbamalu G, Khoo W, Bernstein A, Pawson T: The Fps/Fes protein-tyrosine kinase promotes angiogenesis in transgenic mice, Mol Cell Biol 1994, 14:6755-6763 52. Haigh J, McVeigh J, Greer P: The fps/fes tyrosine kinase is expressed in myeloid, vascular endothelial, epithelial, and neuronal cells and is localized in the trans-golgi network, Cell Growth Differ 1996, 7:931-944 53. Bardelli A, Parsons DW, Silliman N, Ptak J, Szabo S, Saha S, Markowitz S, Willson JK, Parmigiani G, Kinzler KW, Vogelstein B, Velculescu VE: Mutational analysis of the tyrosine kinome in colorectal cancers, Science 2003, 300:949 54. Delfino FJ, Stevenson H, Smithgall TE: A growth-suppressive function for the c-fes protein-tyrosine kinase in colorectal cancer, J Biol Chem 2006, 281:8829-8835 55. Sangrar W, Zirgnibl RA, Gao Y, Muller WJ, Jia Z, Greer PA: An identity crisis for fps/fes: oncogene or tumor suppressor? Cancer Res 2005, 65:3518-3522 56. Feldman RA, Wang E, Hanafusa H: Cytoplasmic localization of the transforming protein of Fujinami sarcoma virus: salt-sensitive association with subcellular components, J Virol 1983, 45:782-791 57. Hao QL, Ferris DK, White G, Heisterkamp N, Groffen J: Nuclear and cytoplasmic location of the FER tyrosine kinase, Mol Cell Biol 1991, 11:1180-1183 58. Ben-Dor I, Bern O, Tennenbaum T, Nir U: Cell cycle-dependent nuclear accumulation of the p94fer tyrosine kinase is regulated by its NH2 terminus and is affected by kinase domain integrity and ATP binding, Cell Growth Differ 1999, 10:113-129 59. Tagliafico E, Siena M, Zanocco-Marani T, Manfredini R, Tenedini E, Montanari M, Grande A, Ferrari S: Requirement of the coiled-coil domains of p92(c-Fes) for nuclear localization in myeloid cells upon induction of differentiation, Oncogene 2003, 22:1712-1723 60. Yates KE, Lynch MR, Wong SG, Slamon DJ, Gasson JC: Human c-FES is a nuclear tyrosine kinase, Oncogene 1995, 10:1239-1242 61. Schwartz Y, Ben-Dor I, Navon A, Motro B, Nir U: Tyrosine phosphorylation of the TATA element modulatory factor by the FER nuclear tyrosine kinases, FEBS Lett 1998, 434:339-345 62. Craig AW, Zirngibl R, Greer P: Disruption of coiled-coil domains in Fer protein-tyrosine kinase abolishes trimerization but not kinase activation, J Biol Chem 1999, 274:19934-19942 63. Zirngibl R, Schulze D, Mirski SE, Cole SP, Greer PA: Subcellular localization analysis of the closely related Fps/Fes and Fer protein-tyrosine kinases suggests a distinct role for Fps/Fes in vesicular trafficking, Exp Cell Res 2001, 266:87-94 64. Huynh H, Bottini N, Williams S, Cherepanov V, Musumeci L, Saito K, Bruckner S, Vachon E, Wang X, Kruger J, Chow CW, Pellecchia M, Monosov E, Greer PA, Trimble W, Downey GP, Mustelin T: Control of vesicle fusion by a tyrosine phosphatase, Nat Cell Biol 2004, 6:831-839 65. Robinson DR, Wu YM, Lin SF: The protein tyrosine kinase family of the human genome, Oncogene 2000, 19:5548-5557 66. Sadowski I, Stone JC, Pawson T: A noncatalytic domain conserved among cytoplasmic protein-tyrosine kinases modifies the kinase function and transforming activity of Fujinami sarcoma virus P130gag-fps, Mol Cell Biol 1986, 6:4396-4408

157

67. Tsujita K, Suetsugu S, Sasaki N, Furutani M, Oikawa T, Takenawa T: Coordination between the actin cytoskeleton and membrane deformation by a novel membrane tubulation domain of PCH proteins is involved in endocytosis, J Cell Biol 2006, 172:269-279 68. Aspenstrom P: A Cdc42 target protein with homology to the non-kinase domain of FER has a potential role in regulating the actin cytoskeleton, Curr Biol 1997, 7:479-487 69. Yeung YG, Soldera S, Stanley ER: A novel macrophage actin-associated protein (MAYP) is tyrosine-phosphorylated following colony stimulating factor-1 stimulation, J Biol Chem 1998, 273:30638-30642 70. Modregger J, Ritter B, Witter B, Paulsson M, Plomann M: All three PACSIN isoforms bind to endocytic proteins and inhibit endocytosis, J Cell Sci 2000, 113 Pt 24:4511-4521 71. Tian L, Nelson DL, Stewart DM: Cdc42-interacting protein 4 mediates binding of the Wiskott-Aldrich syndrome protein to microtubules, J Biol Chem 2000, 275:7854-7861 72. Lippincott J, Li R: Involvement of PCH family proteins in cytokinesis and actin distribution, Microsc Res Tech 2000, 49:168-172 73. Frost A, De Camilli P, Unger VM: F-BAR proteins join the BAR family fold, Structure 2007, 15:751-753 74. Itoh T, De Camilli P: BAR, F-BAR (EFC) and ENTH/ANTH domains in the regulation of membrane-cytosol interfaces and membrane curvature, Biochim Biophys Acta 2006, 1761:897-912 75. Itoh T, Erdmann KS, Roux A, Habermann B, Werner H, De Camilli P: Dynamin and the actin cytoskeleton cooperatively regulate plasma membrane invagination by BAR and F-BAR proteins, Dev Cell 2005, 9:791-804 76. Dawson JC, Legg JA, Machesky LM: Bar domain proteins: a role in tubulation, scission and actin assembly in clathrin-mediated endocytosis, Trends Cell Biol 2006, 16:493-498 77. Yang L, Wang L, Zheng Y: Gene targeting of Cdc42 and Cdc42GAP affirms the critical involvement of Cdc42 in filopodia induction, directed migration, and proliferation in primary mouse embryonic fibroblasts, Mol Biol Cell 2006, 17:4675-4685 78. Richnau N, Aspenstrom P: Rich, a rho GTPase-activating protein domain-containing protein involved in signaling by Cdc42 and Rac1, J Biol Chem 2001, 276:35060-35070 79. Linder S, Hufner K, Wintergerst U, Aepfelbacher M: Microtubule-dependent formation of podosomal adhesion structures in primary human macrophages, J Cell Sci 2000, 113 Pt 23:4165-4176 80. Shimada A, Niwa H, Tsujita K, Suetsugu S, Nitta K, Hanawa-Suetsugu K, Akasaka R, Nishino Y, Toyama M, Chen L, Liu ZJ, Wang BC, Yamamoto M, Terada T, Miyazawa A, Tanaka A, Sugano S, Shirouzu M, Nagayama K, Takenawa T, Yokoyama S: Curved EFC/F-BAR-domain dimers are joined end to end into a filament for membrane invagination in endocytosis, Cell 2007, 129:761-772 81. Peter BJ, Kent HM, Mills IG, Vallis Y, Butler PJ, Evans PR, McMahon HT: BAR domains as sensors of membrane curvature: the amphiphysin BAR structure, Science 2004, 303:495-499 82. Takahashi S, Inatome R, Hotta A, Qin Q, Hackenmiller R, Simon MC, Yamamura H, Yanagi S: Role for Fes/Fps tyrosine kinase in microtubule nucleation through is Fes/CIP4 homology domain, J Biol Chem 2003, 278:49129-49133

158

83. Laurent CE, Delfino FJ, Cheng HY, Smithgall TE: The human c-Fes tyrosine kinase binds tubulin and microtubules through separate domains and promotes microtubule assembly, Mol Cell Biol 2004, 24:9351-9358 84. Liu J, Zheng Q, Deng Y, Cheng CS, Kallenbach NR, Lu M: A seven-helix coiled coil, Proc Natl Acad Sci U S A 2006, 103:15457-15462 85. Read RD, Lionberger JM, Smithgall TE: Oligomerization of the Fes tyrosine kinase. Evidence for a coiled-coil domain in the unique N-terminal region, J Biol Chem 1997, 272:18498-18503 86. Cheng H, Rogers JA, Dunham NA, Smithgall TE: Regulation of c-Fes tyrosine kinase and biological activities by N-terminal coiled-coil oligomerization domains, Mol Cell Biol 1999, 19:8335-8343 87. Smithgall TE, Rogers JA, Peters KL, Li J, Briggs SD, Lionberger JM, Cheng H, Shibata A, Scholtz B, Schreiner S, Dunham N: The c-Fes family of protein-tyrosine kinases, Crit Rev Oncog 1998, 9:43-62 88. DeClue JE, Sadowski I, Martin GS, Pawson T: A conserved domain regulates interactions of the v-fps protein-tyrosine kinase with the host cell, Proc Natl Acad Sci U S A 1987, 84:9064-9068 89. Pawson T: Tyrosine kinase signalling pathways, Princess Takamatsu Symp 1994, 24:303-322 90. Stone JC, Pawson T: Correspondence between immunological and functional domains in the transforming protein of Fujinami sarcoma virus, J Virol 1985, 55:721-727 91. Hjermstad SJ, Peters KL, Briggs SD, Glazer RI, Smithgall TE: Regulation of the human c-fes protein tyrosine kinase (p93c-fes) by its src homology 2 domain and major autophosphorylation site (Tyr-713), Oncogene 1993, 8:2283-2292 92. Songyang Z, Shoelson SE, McGlade J, Olivier P, Pawson T, Bustelo XR, Barbacid M, Sabe H, Hanafusa H, Yi T, et al.: Specific motifs recognized by the SH2 domains of Csk, 3BP2, fps/fes, GRB-2, HCP, SHC, Syk, and Vav, Mol Cell Biol 1994, 14:2777-2785 93. Songyang Z, Carraway KL, 3rd, Eck MJ, Harrison SC, Feldman RA, Mohammadi M, Schlessinger J, Hubbard SR, Smith DP, Eng C, et al.: Catalytic specificity of protein-tyrosine kinases is critical for selective signalling, Nature 1995, 373:536-539 94. Fang F, Ahmad S, Lei J, Klecker RW, Trepel JB, Smithgall TE, Glazer RI: Effect of the mutation of tyrosine 713 in p93c-fes on its catalytic activity and ability to promote myeloid differentiation in K562 cells, Biochemistry 1993, 32:6995-7001 95. Weinmaster G, Zoller MJ, Smith M, Hinze E, Pawson T: Mutagenesis of Fujinami sarcoma virus: evidence that tyrosine phosphorylation of P130gag-fps modulates its biological activity, Cell 1984, 37:559-568 96. Rogers JA, Read RD, Li J, Peters KL, Smithgall TE: Autophosphorylation of the Fes tyrosine kinase. Evidence for an intermolecular mechanism involving two kinase domain tyrosine residues, J Biol Chem 1996, 271:17519-17525 97. Udell CM, Samayawardhena LA, Kawakami Y, Kawakami T, Craig AW: Fer and Fps/Fes participate in a Lyn-dependent pathway from FcepsilonRI to platelet-endothelial cell adhesion molecule 1 to limit mast cell activation, J Biol Chem 2006, 281:20949-20957 98. Takashima Y, Delfino FJ, Engen JR, Superti-Furga G, Smithgall TE: Regulation of c-Fes tyrosine kinase activity by coiled-coil and SH2 domains: analysis with Saccharomyces cerevisiae, Biochemistry 2003, 42:3567-3574 99. Sangrar W, Mewburn JD, Vincent SG, Fisher JT, Greer PA: Vascular defects in gain-of-function fps/fes transgenic mice correlate with PDGF- and VEGF-induced activation of mutant Fps/Fes kinase in endothelial cells, J Thromb Haemost 2004, 2:820-832

159

100. Wang SJ, Greer P, Auerbach R: Isolation and propagation of yolk-sac-derived endothelial cells from a hypervascular transgenic mouse expressing a gain-of-function fps/fes proto-oncogene, In Vitro Cell Dev Biol Anim 1996, 32:292-299 101. Senis Y, Zirngibl R, McVeigh J, Haman A, Hoang T, Greer PA: Targeted disruption of the murine fps/fes proto-oncogene reveals that Fps/Fes kinase activity is dispensable for hematopoiesis, Mol Cell Biol 1999, 19:7436-7446 102. Parsons SA, Greer PA: The Fps/Fes kinase regulates the inflammatory response to endotoxin through down-regulation of TLR4, NF-kappaB activation, and TNF-alpha secretion in macrophages, J Leukoc Biol 2006, 80:1522-1528 103. Parsons SA, Mewburn JD, Truesdell P, Greer PA: The Fps/Fes kinase regulates leucocyte recruitment and extravasation during inflammation, Immunology 2007, 122:542-550 104. Hackenmiller R, Kim J, Feldman RA, Simon MC: Abnormal Stat activation, hematopoietic homeostasis, and innate immunity in c-fes-/- mice, Immunity 2000, 13:397-407 105. Thomas G: Furin at the cutting edge: from protein traffic to embryogenesis and disease, Nat Rev Mol Cell Biol 2002, 3:753-766 106. Roebroek AJ, Umans L, Pauli IG, Robertson EJ, van Leuven F, Van de Ven WJ, Constam DB: Failure of ventral closure and axial rotation in embryos lacking the proprotein convertase Furin, Development 1998, 125:4863-4876 107. Chodosh LA, Gardner HP, Rajan JV, Stairs DB, Marquis ST, Leder PA: Protein kinase expression during murine mammary development, Dev Biol 2000, 219:259-276 108. Guy CT, Cardiff RD, Muller WJ: Induction of mammary tumors by expression of polyomavirus middle T oncogene: a transgenic mouse model for metastatic disease, Mol Cell Biol 1992, 12:954-961 109. Richert MM, Schwertfeger KL, Ryder JW, Anderson SM: An atlas of mouse mammary gland development, J Mammary Gland Biol Neoplasia 2000, 5:227-241 110. Lochter A: Plasticity of mammary epithelia during normal development and neoplastic progression, Biochem Cell Biol 1998, 76:997-1008 111. Williams JM, Daniel CW: Mammary ductal elongation: differentiation of myoepithelium and basal lamina during branching morphogenesis, Dev Biol 1983, 97:274-290 112. Emerman JT, Vogl AW: Cell size and shape changes in the myoepithelium of the mammary gland during differentiation, Anat Rec 1986, 216:405-415 113. Wiseman BS, Werb Z: Stromal effects on mammary gland development and breast cancer, Science 2002, 296:1046-1049 114. Streuli CH, Edwards GM: Control of normal mammary epithelial phenotype by integrins, J Mammary Gland Biol Neoplasia 1998, 3:151-163 115. Streuli CH, Bissell MJ: Expression of extracellular matrix components is regulated by substratum, J Cell Biol 1990, 110:1405-1415 116. Robinson GW, Karpf AB, Kratochwil K: Regulation of mammary gland development by tissue interaction, J Mammary Gland Biol Neoplasia 1999, 4:9-19 117. Sakakura T, Sakagami Y, Nishizuka Y: Dual origin of mesenchymal tissues participating in mouse mammary gland embryogenesis, Dev Biol 1982, 91:202-207 118. Hovey RC, Trott JF: Morphogenesis of mammary gland development, Adv Exp Med Biol 2004, 554:219-228

160

119. Hovey RC, Trott JF, Vonderhaar BK: Establishing a framework for the functional mammary gland: from endocrinology to morphology, J Mammary Gland Biol Neoplasia 2002, 7:17-38 120. Bocchinfuso WP, Lindzey JK, Hewitt SC, Clark JA, Myers PH, Cooper R, Korach KS: Induction of mammary gland development in estrogen receptor-alpha knockout mice, Endocrinology 2000, 141:2982-2994 121. Daniel CW, Robinson S, Silberstein GB: The transforming growth factors beta in development and functional differentiation of the mouse mammary gland, Adv Exp Med Biol 2001, 501:61-70 122. Gallego MI, Binart N, Robinson GW, Okagaki R, Coschigano KT, Perry J, Kopchick JJ, Oka T, Kelly PA, Hennighausen L: Prolactin, growth hormone, and epidermal growth factor activate Stat5 in different compartments of mammary tissue and exert different and overlapping developmental effects, Dev Biol 2001, 229:163-175 123. Ruan W, Kleinberg DL: Insulin-like growth factor I is essential for terminal end bud formation and ductal morphogenesis during mammary development, Endocrinology 1999, 140:5075-5081 124. Korach KS: Insights from the study of animals lacking functional estrogen receptor, Science 1994, 266:1524-1527 125. Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O: Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene, Proc Natl Acad Sci U S A 1993, 90:11162-11166 126. Ciarloni L, Mallepell S, Brisken C: Amphiregulin is an essential mediator of estrogen receptor alpha function in mammary gland development, Proc Natl Acad Sci U S A 2007, 104:5455-5460 127. Kleinberg DL: Early mammary development: growth hormone and IGF-1, J Mammary Gland Biol Neoplasia 1997, 2:49-57 128. Kleinberg DL, Feldman M, Ruan W: IGF-I: an essential factor in terminal end bud formation and ductal morphogenesis, J Mammary Gland Biol Neoplasia 2000, 5:7-17 129. Horseman ND, Zhao W, Montecino-Rodriguez E, Tanaka M, Nakashima K, Engle SJ, Smith F, Markoff E, Dorshkind K: Defective mammopoiesis, but normal hematopoiesis, in mice with a targeted disruption of the prolactin gene, Embo J 1997, 16:6926-6935 130. Ormandy CJ, Camus A, Barra J, Damotte D, Lucas B, Buteau H, Edery M, Brousse N, Babinet C, Binart N, Kelly PA: Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse, Genes Dev 1997, 11:167-178 131. Ormandy CJ, Naylor M, Harris J, Robertson F, Horseman ND, Lindeman GJ, Visvader J, Kelly PA: Investigation of the transcriptional changes underlying functional defects in the mammary glands of prolactin receptor knockout mice, Recent Prog Horm Res 2003, 58:297-323 132. Reese J, Binart N, Brown N, Ma WG, Paria BC, Das SK, Kelly PA, Dey SK: Implantation and decidualization defects in prolactin receptor (PRLR)-deficient mice are mediated by ovarian but not uterine PRLR, Endocrinology 2000, 141:1872-1881 133. Haslam SZ: The ontogeny of mouse mammary gland responsiveness to ovarian steroid hormones, Endocrinology 1989, 125:2766-2772 134. Wang S, Counterman LJ, Haslam SZ: Progesterone action in normal mouse mammary gland, Endocrinology 1990, 127:2183-2189 135. Andres AC, Strange R: Apoptosis in the estrous and menstrual cycles, J Mammary Gland Biol Neoplasia 1999, 4:221-228

161

136. Robinson GW, McKnight RA, Smith GH, Hennighausen L: Mammary epithelial cells undergo secretory differentiation in cycling virgins but require pregnancy for the establishment of terminal differentiation, Development 1995, 121:2079-2090 137. Harris J, Stanford PM, Sutherland K, Oakes SR, Naylor MJ, Robertson FG, Blazek KD, Kazlauskas M, Hilton HN, Wittlin S, Alexander WS, Lindeman GJ, Visvader JE, Ormandy CJ: Socs2 and elf5 mediate prolactin-induced mammary gland development, Mol Endocrinol 2006, 20:1177-1187 138. Feng Y, Manka D, Wagner KU, Khan SA: Estrogen receptor-alpha expression in the mammary epithelium is required for ductal and alveolar morphogenesis in mice, Proc Natl Acad Sci U S A 2007, 104:14718-14723 139. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA, Jr., Shyamala G, Conneely OM, O'Malley BW: Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities, Genes Dev 1995, 9:2266-2278 140. Ismail PM, Amato P, Soyal SM, DeMayo FJ, Conneely OM, O'Malley BW, Lydon JP: Progesterone involvement in breast development and tumorigenesis--as revealed by progesterone receptor "knockout" and "knockin" mouse models, Steroids 2003, 68:779-787 141. Brisken C, Park S, Vass T, Lydon JP, O'Malley BW, Weinberg RA: A paracrine role for the epithelial progesterone receptor in mammary gland development, Proc Natl Acad Sci U S A 1998, 95:5076-5081 142. Brisken C, Heineman A, Chavarria T, Elenbaas B, Tan J, Dey SK, McMahon JA, McMahon AP, Weinberg RA: Essential function of Wnt-4 in mammary gland development downstream of progesterone signaling, Genes Dev 2000, 14:650-654 143. Brisken C, Rajaram RD: Alveolar and lactogenic differentiation, J Mammary Gland Biol Neoplasia 2006, 11:239-248 144. Anderson SM, Rudolph MC, McManaman JL, Neville MC: Key stages in mammary gland development. Secretory activation in the mammary gland: it's not just about milk protein synthesis! Breast Cancer Res 2007, 9:204 145. Liu X, Robinson GW, Hennighausen L: Activation of Stat5a and Stat5b by tyrosine phosphorylation is tightly linked to mammary gland differentiation, Mol Endocrinol 1996, 10:1496-1506 146. Teglund S, McKay C, Schuetz E, van Deursen JM, Stravopodis D, Wang D, Brown M, Bodner S, Grosveld G, Ihle JN: Stat5a and Stat5b proteins have essential and nonessential, or redundant, roles in cytokine responses, Cell 1998, 93:841-850 147. Sjaastad MD, Zettl KS, Parry G, Firestone GL, Machen TE: Hormonal regulation of the polarized function and distribution of Na/H exchange and Na/HCO3 cotransport in cultured mammary epithelial cells, J Cell Biol 1993, 122:589-600 148. Miyoshi K, Shillingford JM, Smith GH, Grimm SL, Wagner KU, Oka T, Rosen JM, Robinson GW, Hennighausen L: Signal transducer and activator of transcription (Stat) 5 controls the proliferation and differentiation of mammary alveolar epithelium, J Cell Biol 2001, 155:531-542 149. Oakes SR, Hilton HN, Ormandy CJ: The alveolar switch: coordinating the proliferative cues and cell fate decisions that drive the formation of lobuloalveoli from ductal epithelium, Breast Cancer Res 2006, 8:207 150. Zhou J, Chehab R, Tkalcevic J, Naylor MJ, Harris J, Wilson TJ, Tsao S, Tellis I, Zavarsek S, Xu D, Lapinskas EJ, Visvader J, Lindeman GJ, Thomas R, Ormandy CJ, Hertzog PJ, Kola I, Pritchard MA: Elf5 is essential for early embryogenesis and mammary gland development during pregnancy and lactation, Embo J 2005, 24:635-644

162

151. Johansson EM, Kannius-Janson M, Gritli-Linde A, Bjursell G, Nilsson J: Nuclear factor 1-C2 is regulated by prolactin and shows a distinct expression pattern in the mouse mammary epithelial cells during development, Mol Endocrinol 2005, 19:992-1003 152. Nilsson J, Bjursell G, Kannius-Janson M: Nuclear Jak2 and transcription factor NF1-C2: a novel mechanism of prolactin signaling in mammary epithelial cells, Mol Cell Biol 2006, 26:5663-5674 153. Hovey RC, Harris J, Hadsell DL, Lee AV, Ormandy CJ, Vonderhaar BK: Local insulin-like growth factor-II mediates prolactin-induced mammary gland development, Mol Endocrinol 2003, 17:460-471 154. Bradbury JM, Edwards PA, Niemeyer CC, Dale TC: Wnt-4 expression induces a pregnancy-like growth pattern in reconstituted mammary glands in virgin mice, Dev Biol 1995, 170:553-563 155. Luetteke NC, Qiu TH, Fenton SE, Troyer KL, Riedel RF, Chang A, Lee DC: Targeted inactivation of the EGF and amphiregulin genes reveals distinct roles for EGF receptor ligands in mouse mammary gland development, Development 1999, 126:2739-2750 156. Fata JE, Kong YY, Li J, Sasaki T, Irie-Sasaki J, Moorehead RA, Elliott R, Scully S, Voura EB, Lacey DL, Boyle WJ, Khokha R, Penninger JM: The osteoclast differentiation factor osteoprotegerin-ligand is essential for mammary gland development, Cell 2000, 103:41-50 157. Neville MC, McFadden TB, Forsyth I: Hormonal regulation of mammary differentiation and milk secretion, J Mammary Gland Biol Neoplasia 2002, 7:49-66 158. Rosen JM, Wyszomierski SL, Hadsell D: Regulation of milk protein gene expression, Annu Rev Nutr 1999, 19:407-436 159. Baruch A, Shani M, Hurwitz DR, Barash I: Developmental regulation of the ovine beta-lactoglobulin/human serum albumin transgene is distinct from that of the beta-lactoglobulin and the endogenous beta-casein genes in the mammary gland of transgenic mice, Dev Genet 1995, 16:241-252 160. Rudolph MC, McManaman JL, Hunter L, Phang T, Neville MC: Functional development of the mammary gland: use of expression profiling and trajectory clustering to reveal changes in gene expression during pregnancy, lactation, and involution, J Mammary Gland Biol Neoplasia 2003, 8:287-307 161. Stacey A, Schnieke A, Kerr M, Scott A, McKee C, Cottingham I, Binas B, Wilde C, Colman A: Lactation is disrupted by alpha-lactalbumin deficiency and can be restored by human alpha-lactalbumin gene replacement in mice, Proc Natl Acad Sci U S A 1995, 92:2835-2839 162. Camps M, Vilaro S, Testar X, Palacin M, Zorzano A: High and polarized expression of GLUT1 glucose transporters in epithelial cells from mammary gland: acute down-regulation of GLUT1 carriers by weaning, Endocrinology 1994, 134:924-934 163. Nemeth BA, Tsang SW, Geske RS, Haney PM: Golgi targeting of the GLUT1 glucose transporter in lactating mouse mammary gland, Pediatr Res 2000, 47:444-450 164. Naylor MJ, Oakes SR, Gardiner-Garden M, Harris J, Blazek K, Ho TW, Li FC, Wynick D, Walker AM, Ormandy CJ: Transcriptional changes underlying the secretory activation phase of mammary gland development, Mol Endocrinol 2005, 19:1868-1883 165. Boxer RB, Stairs DB, Dugan KD, Notarfrancesco KL, Portocarrero CP, Keister BA, Belka GK, Cho H, Rathmell JC, Thompson CB, Birnbaum MJ, Chodosh LA: Isoform-specific requirement for Akt1 in the developmental regulation of cellular metabolism during lactation, Cell Metab 2006, 4:475-490

163

166. Schwertfeger KL, McManaman JL, Palmer CA, Neville MC, Anderson SM: Expression of constitutively activated Akt in the mammary gland leads to excess lipid synthesis during pregnancy and lactation, J Lipid Res 2003, 44:1100-1112 167. Berwick DC, Hers I, Heesom KJ, Moule SK, Tavare JM: The identification of ATP-citrate lyase as a protein kinase B (Akt) substrate in primary adipocytes, J Biol Chem 2002, 277:33895-33900 168. Horton JD, Goldstein JL, Brown MS: SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver, J Clin Invest 2002, 109:1125-1131 169. Porstmann T, Griffiths B, Chung YL, Delpuech O, Griffiths JR, Downward J, Schulze A: PKB/Akt induces transcription of enzymes involved in cholesterol and fatty acid biosynthesis via activation of SREBP, Oncogene 2005, 24:6465-6481 170. Sundqvist A, Bengoechea-Alonso MT, Ye X, Lukiyanchuk V, Jin J, Harper JW, Ericsson J: Control of lipid metabolism by phosphorylation-dependent degradation of the SREBP family of transcription factors by SCF(Fbw7), Cell Metab 2005, 1:379-391 171. Cross DA, Alessi DR, Cohen P, Andjelkovich M, Hemmings BA: Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B, Nature 1995, 378:785-789 172. Nguyen DA, Neville MC: Tight junction regulation in the mammary gland, J Mammary Gland Biol Neoplasia 1998, 3:233-246 173. Adriance MC, Inman JL, Petersen OW, Bissell MJ: Myoepithelial cells: good fences make good neighbors, Breast Cancer Res 2005, 7:190-197 174. Faraldo MM, Deugnier MA, Lukashev M, Thiery JP, Glukhova MA: Perturbation of beta1-integrin function alters the development of murine mammary gland, Embo J 1998, 17:2139-2147 175. Mather IH, Keenan TW: Origin and secretion of milk lipids, J Mammary Gland Biol Neoplasia 1998, 3:259-273 176. McManaman JL, Reyland ME, Thrower EC: Secretion and fluid transport mechanisms in the mammary gland: comparisons with the exocrine pancreas and the salivary gland, J Mammary Gland Biol Neoplasia 2006, 11:249-268 177. Neville MC, Morton J, Umemura S: Lactogenesis. The transition from pregnancy to lactation, Pediatr Clin North Am 2001, 48:35-52 178. Wagner KU, Young WS, 3rd, Liu X, Ginns EI, Li M, Furth PA, Hennighausen L: Oxytocin and milk removal are required for post-partum mammary-gland development, Genes Funct 1997, 1:233-244 179. Young WS, 3rd, Shepard E, Amico J, Hennighausen L, Wagner KU, LaMarca ME, McKinney C, Ginns EI: Deficiency in mouse oxytocin prevents milk ejection, but not fertility or parturition, J Neuroendocrinol 1996, 8:847-853 180. Da Costa TH, Taylor K, Ilic V, Williamson DH: Regulation of milk lipid secretion: effects of oxytocin, prolactin and ionomycin on triacylglycerol release from rat mammary gland slices, Biochem J 1995, 308 (Pt 3):975-981 181. Quarrie LH, Addey CV, Wilde CJ: Programmed cell death during mammary tissue involution induced by weaning, litter removal, and milk stasis, J Cell Physiol 1996, 168:559-569 182. Lund LR, Romer J, Thomasset N, Solberg H, Pyke C, Bissell MJ, Dano K, Werb Z: Two distinct phases of apoptosis in mammary gland involution: proteinase-independent and -dependent pathways, Development 1996, 122:181-193 183. Green KA, Naylor MJ, Lowe ET, Wang P, Marshman E, Streuli CH: Caspase-mediated Cleavage of Insulin Receptor Substrate, J Biol Chem 2004, 279:25149-25156

164

184. Marti A, Lazar H, Ritter P, Jaggi R: Transcription factor activities and gene expression during mouse mammary gland involution, J Mammary Gland Biol Neoplasia 1999, 4:145-152 185. Gigliotti AP, DeWille JW: Local signals induce CCAAT/enhancer binding protein-delta (C/EBP-delta) and C/EBP-beta mRNA expression in the involuting mouse mammary gland, Breast Cancer Res Treat 1999, 58:57-63 186. Merlo GR, Basolo F, Fiore L, Duboc L, Hynes NE: p53-dependent and p53-independent activation of apoptosis in mammary epithelial cells reveals a survival function of EGF and insulin, J Cell Biol 1995, 128:1185-1196 187. Faure E, Heisterkamp N, Groffen J, Kaartinen V: Differential expression of TGF-beta isoforms during postlactational mammary gland involution, Cell Tissue Res 2000, 300:89-95 188. Zinser GM, Welsh J: Accelerated mammary gland development during pregnancy and delayed postlactational involution in vitamin D3 receptor null mice, Mol Endocrinol 2004, 18:2208-2223 189. Kritikou EA, Sharkey A, Abell K, Came PJ, Anderson E, Clarkson RW, Watson CJ: A dual, non-redundant, role for LIF as a regulator of development and STAT3-mediated cell death in mammary gland, Development 2003, 130:3459-3468 190. Clarkson RW, Wayland MT, Lee J, Freeman T, Watson CJ: Gene expression profiling of mammary gland development reveals putative roles for death receptors and immune mediators in post-lactational regression, Breast Cancer Res 2004, 6:R92-109 191. Thangaraju M, Rudelius M, Bierie B, Raffeld M, Sharan S, Hennighausen L, Huang AM, Sterneck E: C/EBPdelta is a crucial regulator of pro-apoptotic gene expression during mammary gland involution, Development 2005, 132:4675-4685 192. Abell K, Bilancio A, Clarkson RW, Tiffen PG, Altaparmakov AI, Burdon TG, Asano T, Vanhaesebroeck B, Watson CJ: Stat3-induced apoptosis requires a molecular switch in PI(3)K subunit composition, Nat Cell Biol 2005, 7:392-398 193. Stein T, Salomonis N, Gusterson BA: Mammary gland involution as a multi-step process, J Mammary Gland Biol Neoplasia 2007, 12:25-35 194. Vallorosi CJ, Day KC, Zhao X, Rashid MG, Rubin MA, Johnson KR, Wheelock MJ, Day ML: Truncation of the beta-catenin binding domain of E-cadherin precedes epithelial apoptosis during prostate and mammary involution, J Biol Chem 2000, 275:3328-3334 195. Watson CJ: Involution: apoptosis and tissue remodelling that convert the mammary gland from milk factory to a quiescent organ, Breast Cancer Res 2006, 8:203 196. Stein T, Morris JS, Davies CR, Weber-Hall SJ, Duffy MA, Heath VJ, Bell AK, Ferrier RK, Sandilands GP, Gusterson BA: Involution of the mouse mammary gland is associated with an immune cascade and an acute-phase response, involving LBP, CD14 and STAT3, Breast Cancer Res 2004, 6:R75-91 197. Monks J, Rosner D, Geske FJ, Lehman L, Hanson L, Neville MC, Fadok VA: Epithelial cells as phagocytes: apoptotic epithelial cells are engulfed by mammary alveolar epithelial cells and repress inflammatory mediator release, Cell Death Differ 2005, 12:107-114 198. Baxter FO, Neoh K, Tevendale MC: The beginning of the end: death signaling in early involution, J Mammary Gland Biol Neoplasia 2007, 12:3-13 199. McMahon CD, Farr VC, Singh K, Wheeler TT, Davis SR: Decreased expression of beta1-integrin and focal adhesion kinase in epithelial cells may initiate involution of mammary glands, J Cell Physiol 2004, 200:318-325

165

200. Strange R, Li F, Saurer S, Burkhardt A, Friis RR: Apoptotic cell death and tissue remodelling during mouse mammary gland involution, Development 1992, 115:49-58 201. Fadok VA: Clearance: the last and often forgotten stage of apoptosis, J Mammary Gland Biol Neoplasia 1999, 4:203-211 202. Monks J, Geske FJ, Lehman L, Fadok VA: Do inflammatory cells participate in mammary gland involution? J Mammary Gland Biol Neoplasia 2002, 7:163-176 203. McManaman JL, Neville MC: Mammary physiology and milk secretion, Adv Drug Deliv Rev 2003, 55:629-641 204. West DW, Clegg RA: Casein kinase activity in rat mammary gland Golgi vesicles. Phosphorylation of endogenous caseins, Eur J Biochem 1983, 137:215-220 205. Pauloin A, Tooze SA, Michelutti I, Delpal S, Ollivier-Bousquet M: The majority of clathrin coated vesicles from lactating rabbit mammary gland arises from the secretory pathway, J Cell Sci 1999, 112 (Pt 22):4089-4100 206. Musch A, Cohen D, Kreitzer G, Rodriguez-Boulan E: cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network, Embo J 2001, 20:2171-2179 207. Kroschewski R, Hall A, Mellman I: Cdc42 controls secretory and endocytic transport to the basolateral plasma membrane of MDCK cells, Nat Cell Biol 1999, 1:8-13 208. Lafont F, Burkhardt JK, Simons K: Involvement of microtubule motors in basolateral and apical transport in kidney cells, Nature 1994, 372:801-803 209. Nickerson SC, Smith JJ, Keenan TW: Role of microtubules in milk secretion--action of colchicine on microtubules and exocytosis of secretory vesicles in rat mammary epithelial cells, Cell Tissue Res 1980, 207:361-376 210. Rennison ME, Handel SE, Wilde CJ, Burgoyne RD: Investigation of the role of microtubules in protein secretion from lactating mouse mammary epithelial cells, J Cell Sci 1992, 102 (Pt 2):239-247 211. Amato PA, Loizzi RF: The identification and localization of actin and actin-like filaments in lactating guinea pig mammary gland alveolar cells, Cell Motil 1981, 1:329-347 212. Weber T, Zemelman BV, McNew JA, Westermann B, Gmachl M, Parlati F, Sollner TH, Rothman JE: SNAREpins: minimal machinery for membrane fusion, Cell 1998, 92:759-772 213. Sollner T, Bennett MK, Whiteheart SW, Scheller RH, Rothman JE: A protein assembly-disassembly pathway in vitro that may correspond to sequential steps of synaptic vesicle docking, activation, and fusion, Cell 1993, 75:409-418 214. Jena BP, Cho SJ, Jeremic A, Stromer MH, Abu-Hamdah R: Structure and composition of the fusion pore, Biophys J 2003, 84:1337-1343 215. Cho SJ, Kelly M, Rognlien KT, Cho JA, Horber JK, Jena BP: SNAREs in opposing bilayers interact in a circular array to form conducting pores, Biophys J 2002, 83:2522-2527 216. Jeremic A, Kelly M, Cho JA, Cho SJ, Horber JK, Jena BP: Calcium drives fusion of SNARE-apposed bilayers, Cell Biol Int 2004, 28:19-31 217. Leabu M: Membrane fusion in cells: molecular machinery and mechanisms, J Cell Mol Med 2006, 10:423-427 218. Jena BP: Secretion machinery at the cell plasma membrane, Curr Opin Struct Biol 2007, 17:437-443 219. Wu CC, Howell KE, Neville MC, Yates JR, 3rd, McManaman JL: Proteomics reveal a link between the endoplasmic reticulum and lipid secretory mechanisms in mammary epithelial cells, Electrophoresis 2000, 21:3470-3482

166

220. Tauchi-Sato K, Ozeki S, Houjou T, Taguchi R, Fujimoto T: The surface of lipid droplets is a phospholipid monolayer with a unique Fatty Acid composition, J Biol Chem 2002, 277:44507-44512 221. Stemberger BH, Walsh RM, Patton S: Morphometric evaluation of lipid droplet associations with secretory vesicles, mitochondria and other components in the lactating cell, Cell Tissue Res 1984, 236:471-475 222. Kraemer J, Schmitz F, Drenckhahn D: Cytoplasmic dynein and dynactin as likely candidates for microtubule-dependent apical targeting of pancreatic zymogen granules, Eur J Cell Biol 1999, 78:265-277 223. McManaman JL, Russell TD, Schaack J, Orlicky DJ, Robenek H: Molecular determinants of milk lipid secretion, J Mammary Gland Biol Neoplasia 2007, 12:259-268 224. Vorbach C, Scriven A, Capecchi MR: The housekeeping gene xanthine oxidoreductase is necessary for milk fat droplet enveloping and secretion: gene sharing in the lactating mammary gland, Genes Dev 2002, 16:3223-3235 225. Ogg SL, Weldon AK, Dobbie L, Smith AJ, Mather IH: Expression of butyrophilin (Btn1a1) in lactating mammary gland is essential for the regulated secretion of milk-lipid droplets, Proc Natl Acad Sci U S A 2004, 101:10084-10089 226. Kowluru A, Seavey SE, Rhodes CJ, Metz SA: A novel regulatory mechanism for trimeric GTP-binding proteins in the membrane and secretory granule fractions of human and rodent beta cells, Biochem J 1996, 313 (Pt 1):97-107 227. Takeichi M: Morphogenetic roles of classic cadherins, Curr Opin Cell Biol 1995, 7:619-627 228. Nose A, Nagafuchi A, Takeichi M: Expressed recombinant cadherins mediate cell sorting in model systems, Cell 1988, 54:993-1001 229. Ozawa M, Engel J, Kemler R: Single amino acid substitutions in one Ca2+ binding site of uvomorulin abolish the adhesive function, Cell 1990, 63:1033-1038 230. Nagar B, Overduin M, Ikura M, Rini JM: Structural basis of calcium-induced E-cadherin rigidification and dimerization, Nature 1996, 380:360-364 231. Pokutta S, Herrenknecht K, Kemler R, Engel J: Conformational changes of the recombinant extracellular domain of E-cadherin upon calcium binding, Eur J Biochem 1994, 223:1019-1026 232. Nose A, Takeichi M: A novel cadherin cell adhesion molecule: its expression patterns associated with implantation and organogenesis of mouse embryos, J Cell Biol 1986, 103:2649-2658 233. Klingelhofer J, Laur OY, Troyanovsky RB, Troyanovsky SM: Dynamic interplay between adhesive and lateral E-cadherin dimers, Mol Cell Biol 2002, 22:7449-7458 234. Troyanovsky S: Cadherin dimers in cell-cell adhesion, Eur J Cell Biol 2005, 84:225-233 235. Reynolds AB, Daniel J, McCrea PD, Wheelock MJ, Wu J, Zhang Z: Identification of a new catenin: the tyrosine kinase substrate p120cas associates with E-cadherin complexes, Mol Cell Biol 1994, 14:8333-8342 236. Hulsken J, Behrens J, Birchmeier W: Tumor-suppressor gene products in cell contacts: the cadherin-APC-armadillo connection, Curr Opin Cell Biol 1994, 6:711-716 237. Herrenknecht K, Ozawa M, Eckerskorn C, Lottspeich F, Lenter M, Kemler R: The uvomorulin-anchorage protein alpha catenin is a vinculin homologue, Proc Natl Acad Sci U S A 1991, 88:9156-9160 238. Stappert J, Kemler R: A short core region of E-cadherin is essential for catenin binding and is highly phosphorylated, Cell Adhes Commun 1994, 2:319-327

167

239. Pai LM, Kirkpatrick C, Blanton J, Oda H, Takeichi M, Peifer M: Drosophila alpha-catenin and E-cadherin bind to distinct regions of Drosophila Armadillo, J Biol Chem 1996, 271:32411-32420 240. Huber AH, Stewart DB, Laurents DV, Nelson WJ, Weis WI: The cadherin cytoplasmic domain is unstructured in the absence of beta-catenin. A possible mechanism for regulating cadherin turnover, J Biol Chem 2001, 276:12301-12309 241. Chen YT, Stewart DB, Nelson WJ: Coupling assembly of the E-cadherin/beta-catenin complex to efficient endoplasmic reticulum exit and basal-lateral membrane targeting of E-cadherin in polarized MDCK cells, J Cell Biol 1999, 144:687-699 242. Aberle H, Butz S, Stappert J, Weissig H, Kemler R, Hoschuetzky H: Assembly of the cadherin-catenin complex in vitro with recombinant proteins, J Cell Sci 1994, 107 (Pt 12):3655-3663 243. Rimm DL, Koslov ER, Kebriaei P, Cianci CD, Morrow JS: Alpha 1(E)-catenin is an actin-binding and -bundling protein mediating the attachment of F-actin to the membrane adhesion complex, Proc Natl Acad Sci U S A 1995, 92:8813-8817 244. Drees F, Pokutta S, Yamada S, Nelson WJ, Weis WI: Alpha-catenin is a molecular switch that binds E-cadherin-beta-catenin and regulates actin-filament assembly, Cell 2005, 123:903-915 245. Yamada S, Pokutta S, Drees F, Weis WI, Nelson WJ: Deconstructing the cadherin-catenin-actin complex, Cell 2005, 123:889-901 246. Pokutta S, Weis WI: Structure of the dimerization and beta-catenin-binding region of alpha-catenin, Mol Cell 2000, 5:533-543 247. Pollard TD, Beltzner CC: Structure and function of the Arp2/3 complex, Curr Opin Struct Biol 2002, 12:768-774 248. Weis WI, Nelson WJ: Re-solving the cadherin-catenin-actin conundrum, J Biol Chem 2006, 281:35593-35597 249. Roe S, Koslov ER, Rimm DL: A mutation in alpha-catenin disrupts adhesion in clone A cells without perturbing its actin and beta-catenin binding activity, Cell Adhes Commun 1998, 5:283-296 250. Kemler R, Ozawa M: Uvomorulin-catenin complex: cytoplasmic anchorage of a Ca2+-dependent cell adhesion molecule, Bioessays 1989, 11:88-91 251. Reynolds AB, Herbert L, Cleveland JL, Berg ST, Gaut JR: p120, a novel substrate of protein tyrosine kinase receptors and of p60v-src, is related to cadherin-binding factors beta-catenin, plakoglobin and armadillo, Oncogene 1992, 7:2439-2445 252. Thoreson MA, Reynolds AB: Altered expression of the catenin p120 in human cancer: implications for tumor progression, Differentiation 2002, 70:583-589 253. Chen X, Kojima S, Borisy GG, Green KJ: p120 catenin associates with kinesin and facilitates the transport of cadherin-catenin complexes to intercellular junctions, J Cell Biol 2003, 163:547-557 254. Ireton RC, Davis MA, van Hengel J, Mariner DJ, Barnes K, Thoreson MA, Anastasiadis PZ, Matrisian L, Bundy LM, Sealy L, Gilbert B, van Roy F, Reynolds AB: A novel role for p120 catenin in E-cadherin function, J Cell Biol 2002, 159:465-476 255. Davis MA, Ireton RC, Reynolds AB: A core function for p120-catenin in cadherin turnover, J Cell Biol 2003, 163:525-534 256. Nagafuchi A, Takeichi M, Tsukita S: The 102 kd cadherin-associated protein: similarity to vinculin and posttranscriptional regulation of expression, Cell 1991, 65:849-857

168

257. Beavon IR: The E-cadherin-catenin complex in tumour metastasis: structure, function and regulation, Eur J Cancer 2000, 36:1607-1620 258. Larue L, Ohsugi M, Hirchenhain J, Kemler R: E-cadherin null mutant embryos fail to form a trophectoderm epithelium, Proc Natl Acad Sci U S A 1994, 91:8263-8267 259. Haegel H, Larue L, Ohsugi M, Fedorov L, Herrenknecht K, Kemler R: Lack of beta-catenin affects mouse development at gastrulation, Development 1995, 121:3529-3537 260. Torres M, Stoykova A, Huber O, Chowdhury K, Bonaldo P, Mansouri A, Butz S, Kemler R, Gruss P: An alpha-E-catenin gene trap mutation defines its function in preimplantation development, Proc Natl Acad Sci U S A 1997, 94:901-906 261. Reynolds AB: p120-catenin: Past and present, Biochim Biophys Acta 2007, 1773:2-7 262. Boussadia O, Kutsch S, Hierholzer A, Delmas V, Kemler R: E-cadherin is a survival factor for the lactating mouse mammary gland, Mech Dev 2002, 115:53-62 263. Nemade RV, Bierie B, Nozawa M, Bry C, Smith GH, Vasioukhin V, Fuchs E, Hennighausen L: Biogenesis and function of mouse mammary epithelium depends on the presence of functional alpha-catenin, Mech Dev 2004, 121:91-99 264. Cattelino A, Liebner S, Gallini R, Zanetti A, Balconi G, Corsi A, Bianco P, Wolburg H, Moore R, Oreda B, Kemler R, Dejana E: The conditional inactivation of the beta-catenin gene in endothelial cells causes a defective vascular pattern and increased vascular fragility, J Cell Biol 2003, 162:1111-1122 265. Tan X, Behari J, Cieply B, Michalopoulos GK, Monga SP: Conditional deletion of beta-catenin reveals its role in liver growth and regeneration, Gastroenterology 2006, 131:1561-1572 266. Huelsken J, Vogel R, Erdmann B, Cotsarelis G, Birchmeier W: beta-Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin, Cell 2001, 105:533-545 267. Brault V, Moore R, Kutsch S, Ishibashi M, Rowitch DH, McMahon AP, Sommer L, Boussadia O, Kemler R: Inactivation of the beta-catenin gene by Wnt1-Cre-mediated deletion results in dramatic brain malformation and failure of craniofacial development, Development 2001, 128:1253-1264 268. Clevers H: Wnt/beta-catenin signaling in development and disease, Cell 2006, 127:469-480 269. Arce L, Yokoyama NN, Waterman ML: Diversity of LEF/TCF action in development and disease, Oncogene 2006, 25:7492-7504 270. Sheikh F, Chen Y, Liang X, Hirschy A, Stenbit AE, Gu Y, Dalton ND, Yajima T, Lu Y, Knowlton KU, Peterson KL, Perriard JC, Chen J: alpha-E-catenin inactivation disrupts the cardiomyocyte adherens junction, resulting in cardiomyopathy and susceptibility to wall rupture, Circulation 2006, 114:1046-1055 271. Vasioukhin V, Bauer C, Degenstein L, Wise B, Fuchs E: Hyperproliferation and defects in epithelial polarity upon conditional ablation of alpha-catenin in skin, Cell 2001, 104:605-617 272. Kobielak A, Fuchs E: Links between alpha-catenin, NF-kappaB, and squamous cell carcinoma in skin, Proc Natl Acad Sci U S A 2006, 103:2322-2327 273. Perez-Moreno M, Davis MA, Wong E, Pasolli HA, Reynolds AB, Fuchs E: p120-catenin mediates inflammatory responses in the skin, Cell 2006, 124:631-644 274. Davies JA: Do different branching epithelia use a conserved developmental mechanism? Bioessays 2002, 24:937-948

169

275. Davis MA, Reynolds AB: Blocked acinar development, E-cadherin reduction, and intraepithelial neoplasia upon ablation of p120-catenin in the mouse salivary gland, Dev Cell 2006, 10:21-31 276. Masterson J, O'Dea S: Posttranslational truncation of E-cadherin and significance for tumour progression, Cells Tissues Organs 2007, 185:175-179 277. Berx G, Cleton-Jansen AM, Strumane K, de Leeuw WJ, Nollet F, van Roy F, Cornelisse C: E-cadherin is inactivated in a majority of invasive human lobular breast cancers by truncation mutations throughout its extracellular domain, Oncogene 1996, 13:1919-1925 278. Goldstein NS: Does the level of E-cadherin expression correlate with the primary breast carcinoma infiltration pattern and type of systemic metastases? Am J Clin Pathol 2002, 118:425-434 279. Batistatou A, Peschos D, Tsanou H, Charalabopoulos A, Nakanishi Y, Hirohashi S, Agnantis NJ, Charalabopoulos K: In breast carcinoma dysadherin expression is correlated with invasiveness but not with E-cadherin, Br J Cancer 2007, 96:1404-1408 280. Strathdee G: Epigenetic versus genetic alterations in the inactivation of E-cadherin, Semin Cancer Biol 2002, 12:373-379 281. Nass SJ, Herman JG, Gabrielson E, Iversen PW, Parl FF, Davidson NE, Graff JR: Aberrant methylation of the estrogen receptor and E-cadherin 5' CpG islands increases with malignant progression in human breast cancer, Cancer Res 2000, 60:4346-4348 282. Droufakou S, Deshmane V, Roylance R, Hanby A, Tomlinson I, Hart IR: Multiple ways of silencing E-cadherin gene expression in lobular carcinoma of the breast, Int J Cancer 2001, 92:404-408 283. Masciari S, Larsson N, Senz J, Boyd N, Kaurah P, Kandel MJ, Harris LN, Pinheiro HC, Troussard A, Miron P, Tung N, Oliveira C, Collins L, Schnitt S, Garber JE, Huntsman D: Germline E-cadherin mutations in familial lobular breast cancer, J Med Genet 2007, 44:726-731 284. Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, del Barrio MG, Portillo F, Nieto MA: The transcription factor snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression, Nat Cell Biol 2000, 2:76-83 285. Fujita Y, Krause G, Scheffner M, Zechner D, Leddy HE, Behrens J, Sommer T, Birchmeier W: Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the E-cadherin complex, Nat Cell Biol 2002, 4:222-231 286. Sarrio D, Perez-Mies B, Hardisson D, Moreno-Bueno G, Suarez A, Cano A, Martin-Perez J, Gamallo C, Palacios J: Cytoplasmic localization of p120ctn and E-cadherin loss characterize lobular breast carcinoma from preinvasive to metastatic lesions, Oncogene 2004, 23:3272-3283 287. Ilyas M, Novelli M, Wilkinson K, Tomlinson IP, Abbasi AM, Forbes A, Talbot IC: Tumour recurrence is associated with Jass grouping but not with differences in E-cadherin expression in moderately differentiated Dukes' B colorectal cancers, J Clin Pathol 1997, 50:218-222 288. Mastracci TL, Tjan S, Bane AL, O'Malley FP, Andrulis IL: E-cadherin alterations in atypical lobular hyperplasia and lobular carcinoma in situ of the breast, Mod Pathol 2005, 18:741-751 289. Jeschke U, Mylonas I, Kuhn C, Shabani N, Kunert-Keil C, Schindlbeck C, Gerber B, Friese K: Expression of E-cadherin in human ductal breast cancer carcinoma in situ, invasive carcinomas, their lymph node metastases, their distant metastases, carcinomas with recurrence and in recurrence, Anticancer Res 2007, 27:1969-1974

170

290. Miyoshi K, Hennighausen L: Beta-catenin: a transforming actor on many stages, Breast Cancer Res 2003, 5:63-68 291. Polakis P: Wnt signaling and cancer, Genes Dev 2000, 14:1837-1851 292. Aberle H, Bauer A, Stappert J, Kispert A, Kemler R: beta-catenin is a target for the ubiquitin-proteasome pathway, Embo J 1997, 16:3797-3804 293. Nelson WJ, Nusse R: Convergence of Wnt, beta-catenin, and cadherin pathways, Science 2004, 303:1483-1487 294. Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D, Grosschedl R, Birchmeier W: Functional interaction of beta-catenin with the transcription factor LEF-1, Nature 1996, 382:638-642 295. Polakis P: The many ways of Wnt in cancer, Curr Opin Genet Dev 2007, 17:45-51 296. Webster MT, Rozycka M, Sara E, Davis E, Smalley M, Young N, Dale TC, Wooster R: Sequence variants of the axin gene in breast, colon, and other cancers: an analysis of mutations that interfere with GSK3 binding, Genes Chromosomes Cancer 2000, 28:443-453 297. Dulaimi E, Uzzo RG, Greenberg RE, Al-Saleem T, Cairns P: Detection of bladder cancer in urine by a tumor suppressor gene hypermethylation panel, Clin Cancer Res 2004, 10:1887-1893 298. Syrigos KN, Harrington K, Waxman J, Krausz T, Pignatelli M: Altered gamma-catenin expression correlates with poor survival in patients with bladder cancer, J Urol 1998, 160:1889-1893 299. Dabbs DJ, Bhargava R, Chivukula M: Lobular versus ductal breast neoplasms: the diagnostic utility of p120 catenin, Am J Surg Pathol 2007, 31:427-437 300. Daniel JM, Spring CM, Crawford HC, Reynolds AB, Baig A: The p120(ctn)-binding partner Kaiso is a bi-modal DNA-binding protein that recognizes both a sequence-specific consensus and methylated CpG dinucleotides, Nucleic Acids Res 2002, 30:2911-2919 301. Kelly KF, Spring CM, Otchere AA, Daniel JM: NLS-dependent nuclear localization of p120ctn is necessary to relieve Kaiso-mediated transcriptional repression, J Cell Sci 2004, 117:2675-2686 302. Jiang WG, Davies G, Martin TA, Parr C, Watkins G, Mason MD, Mokbel K, Mansel RE: Targeting matrilysin and its impact on tumor growth in vivo: the potential implications in breast cancer therapy, Clin Cancer Res 2005, 11:6012-6019 303. Benjamin JM, Nelson WJ: Bench to bedside and back again: Molecular mechanisms of alpha-catenin function and roles in tumorigenesis, Semin Cancer Biol 2008, 18:53-64 304. Morton RA, Ewing CM, Nagafuchi A, Tsukita S, Isaacs WB: Reduction of E-cadherin levels and deletion of the alpha-catenin gene in human prostate cancer cells, Cancer Res 1993, 53:3585-3590 305. Pierceall WE, Woodard AS, Morrow JS, Rimm D, Fearon ER: Frequent alterations in E-cadherin and alpha- and beta-catenin expression in human breast cancer cell lines, Oncogene 1995, 11:1319-1326 306. Liu TX, Becker MW, Jelinek J, Wu WS, Deng M, Mikhalkevich N, Hsu K, Bloomfield CD, Stone RM, DeAngelo DJ, Galinsky IA, Issa JP, Clarke MF, Look AT: Chromosome 5q deletion and epigenetic suppression of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation, Nat Med 2007, 13:78-83 307. Matsui S, Shiozaki H, Inoue M, Tamura S, Doki Y, Kadowaki T, Iwazawa T, Shimaya K, Nagafuchi A, Tsukita S, et al.: Immunohistochemical evaluation of alpha-catenin expression in human gastric cancer, Virchows Arch 1994, 424:375-381

171

308. Nakopoulou L, Gakiopoulou-Givalou H, Karayiannakis AJ, Giannopoulou I, Keramopoulos A, Davaris P, Pignatelli M: Abnormal alpha-catenin expression in invasive breast cancer correlates with poor patient survival, Histopathology 2002, 40:536-546 309. Setoyama T, Natsugoe S, Okumura H, Matsumoto M, Uchikado Y, Yokomakura N, Ishigami S, Aikou T: alpha-catenin is a significant prognostic factor than E-cadherin in esophageal squamous cell carcinoma, J Surg Oncol 2007, 95:148-155 310. Mialhe A, Louis J, Montlevier S, Peoch M, Pasquier D, Bosson JL, Rambeaud JJ, Seigneurin D: Expression of E-cadherin and alpha-,beta- and gamma-catenins in human bladder carcinomas: are they good prognostic factors? Invasion Metastasis 1997, 17:124-137 311. Kim YT, Choi EK, Kim JW, Kim DK, Kim SH, Yang WI: Expression of E-cadherin and alpha-, beta-, gamma-catenin proteins in endometrial carcinoma, Yonsei Med J 2002, 43:701-711 312. Kashibuchi K, Tomita K, Schalken JA, Kume H, Takeuchi T, Kitamura T: The prognostic value of E-cadherin, alpha-, beta- and gamma-catenin in bladder cancer patients who underwent radical cystectomy, Int J Urol 2007, 14:789-794 313. Lien WH, Klezovitch O, Fernandez TE, Delrow J, Vasioukhin V: alphaE-catenin controls cerebral cortical size by regulating the hedgehog signaling pathway, Science 2006, 311:1609-1612 314. Giannini AL, Vivanco M, Kypta RM: alpha-catenin inhibits beta-catenin signaling by preventing formation of a beta-catenin*T-cell factor*DNA complex, J Biol Chem 2000, 275:21883-21888 315. Hwang SG, Yu SS, Ryu JH, Jeon HB, Yoo YJ, Eom SH, Chun JS: Regulation of beta-catenin signaling and maintenance of chondrocyte differentiation by ubiquitin-independent proteasomal degradation of alpha-catenin, J Biol Chem 2005, 280:12758-12765 316. Pokutta S, Weis WI: The cytoplasmic face of cell contact sites, Curr Opin Struct Biol 2002, 12:255-262 317. Alema S, Salvatore AM: p120 catenin and phosphorylation: Mechanisms and traits of an unresolved issue, Biochim Biophys Acta 2007, 1773:47-58 318. Danilkovitch-Miagkova A, Leonard EJ: Cross-talk between RON receptor tyrosine kinase and other transmembrane receptors, Histol Histopathol 2001, 16:623-631 319. Moon HS, Choi EA, Park HY, Choi JY, Chung HW, Kim JI, Park WI: Expression and tyrosine phosphorylation of E-cadherin, beta- and gamma-catenin, and epidermal growth factor receptor in cervical cancer cells, Gynecol Oncol 2001, 81:355-359 320. Piedra J, Miravet S, Castano J, Palmer HG, Heisterkamp N, Garcia de Herreros A, Dunach M: p120 Catenin-associated Fer and Fyn tyrosine kinases regulate beta-catenin Tyr-142 phosphorylation and beta-catenin-alpha-catenin Interaction, Mol Cell Biol 2003, 23:2287-2297 321. Xu G, Craig AW, Greer P, Miller M, Anastasiadis PZ, Lilien J, Balsamo J: Continuous association of cadherin with beta-catenin requires the non-receptor tyrosine-kinase Fer, J Cell Sci 2004, 117:3207-3219 322. Atkinson KJ, Rao RK: Role of protein tyrosine phosphorylation in acetaldehyde-induced disruption of epithelial tight junctions, Am J Physiol Gastrointest Liver Physiol 2001, 280:G1280-1288 323. Matteucci E, Ridolfi E, Desiderio MA: Hepatocyte growth factor differently influences Met-E-cadherin phosphorylation and downstream signaling pathway in two models of breast cells, Cell Mol Life Sci 2006, 63:2016-2026

172

324. Burks J, Agazie YM: Modulation of alpha-catenin Tyr phosphorylation by SHP2 positively effects cell transformation induced by the constitutively active FGFR3, Oncogene 2006, 25:7166-7179 325. Lilien J, Balsamo J: The regulation of cadherin-mediated adhesion by tyrosine phosphorylation/dephosphorylation of beta-catenin, Curr Opin Cell Biol 2005, 17:459-465 326. Ozawa M, Kemler R: Altered cell adhesion activity by pervanadate due to the dissociation of alpha-catenin from the E-cadherin.catenin complex, J Biol Chem 1998, 273:6166-6170 327. Kapus A, Di Ciano C, Sun J, Zhan X, Kim L, Wong TW, Rotstein OD: Cell volume-dependent phosphorylation of proteins of the cortical cytoskeleton and cell-cell contact sites. The role of Fyn and FER kinases, J Biol Chem 2000, 275:32289-32298 328. Roura S, Miravet S, Piedra J, Garcia de Herreros A, Dunach M: Regulation of E-cadherin/Catenin association by tyrosine phosphorylation, J Biol Chem 1999, 274:36734-36740 329. Rhee J, Buchan T, Zukerberg L, Lilien J, Balsamo J: Cables links Robo-bound Abl kinase to N-cadherin-bound beta-catenin to mediate Slit-induced modulation of adhesion and transcription, Nat Cell Biol 2007, 9:883-892 330. Bonvini P, An WG, Rosolen A, Nguyen P, Trepel J, Garcia de Herreros A, Dunach M, Neckers LM: Geldanamycin abrogates ErbB2 association with proteasome-resistant beta-catenin in melanoma cells, increases beta-catenin-E-cadherin association, and decreases beta-catenin-sensitive transcription, Cancer Res 2001, 61:1671-1677 331. Zeng G, Apte U, Micsenyi A, Bell A, Monga SP: Tyrosine residues 654 and 670 in beta-catenin are crucial in regulation of Met-beta-catenin interactions, Exp Cell Res 2006, 312:3620-3630 332. Balsamo J, Arregui C, Leung T, Lilien J: The nonreceptor protein tyrosine phosphatase PTP1B binds to the cytoplasmic domain of N-cadherin and regulates the cadherin-actin linkage, J Cell Biol 1998, 143:523-532 333. Rhee J, Lilien J, Balsamo J: Essential tyrosine residues for interaction of the non-receptor protein-tyrosine phosphatase PTP1B with N-cadherin, J Biol Chem 2001, 276:6640-6644 334. Rosato R, Veltmaat JM, Groffen J, Heisterkamp N: Involvement of the tyrosine kinase fer in cell adhesion, Mol Cell Biol 1998, 18:5762-5770 335. Mariner DJ, Anastasiadis P, Keilhack H, Bohmer FD, Wang J, Reynolds AB: Identification of Src phosphorylation sites in the catenin p120ctn, J Biol Chem 2001, 276:28006-28013 336. Reynolds AB, Kanner SB, Wang HC, Parsons JT: Stable association of activated pp60src with two tyrosine-phosphorylated cellular proteins, Mol Cell Biol 1989, 9:3951-3958 337. Mariner DJ, Davis MA, Reynolds AB: EGFR signaling to p120-catenin through phosphorylation at Y228, J Cell Sci 2004, 117:1339-1350 338. Calautti E, Cabodi S, Stein PL, Hatzfeld M, Kedersha N, Paolo Dotto G: Tyrosine phosphorylation and src family kinases control keratinocyte cell-cell adhesion, J Cell Biol 1998, 141:1449-1465 339. Ozawa M, Ohkubo T: Tyrosine phosphorylation of p120(ctn) in v-Src transfected L cells depends on its association with E-cadherin and reduces adhesion activity, J Cell Sci 2001, 114:503-512

173

340. Keilhack H, Hellman U, van Hengel J, van Roy F, Godovac-Zimmermann J, Bohmer FD: The protein-tyrosine phosphatase SHP-1 binds to and dephosphorylates p120 catenin, J Biol Chem 2000, 275:26376-26384 341. Castano J, Solanas G, Casagolda D, Raurell I, Villagrasa P, Bustelo XR, Garcia de Herreros A, Dunach M: Specific phosphorylation of p120-catenin regulatory domain differently modulates its binding to RhoA, Mol Cell Biol 2007, 27:1745-1757 342. Noren NK, Liu BP, Burridge K, Kreft B: p120 catenin regulates the actin cytoskeleton via Rho family GTPases, J Cell Biol 2000, 150:567-580 343. Anastasiadis PZ, Moon SY, Thoreson MA, Mariner DJ, Crawford HC, Zheng Y, Reynolds AB: Inhibition of RhoA by p120 catenin, Nat Cell Biol 2000, 2:637-644 344. Anastasiadis PZ, Reynolds AB: Regulation of Rho GTPases by p120-catenin, Curr Opin Cell Biol 2001, 13:604-610 345. Hennighausen L, Robinson GW: Signaling pathways in mammary gland development, Dev Cell 2001, 1:467-475 346. Andrechek ER, Muller WJ: Tyrosine kinase signalling in breast cancer: tyrosine kinase-mediated signal transduction in transgenic mouse models of human breast cancer, Breast Cancer Res 2000, 2:211-216 347. Shapovalova Z, Tabunshchyk K, Greer PA: The Fer tyrosine kinase regulates an axon retraction response to Semaphorin 3A in dorsal root ganglion neurons, BMC Dev Biol 2007, 7:133 348. Shibata A, Laurent CE, Smithgall TE: The c-Fes protein-tyrosine kinase accelerates NGF-induced differentiation of PC12 cells through a PI3K-dependent mechanism, Cell Signal 2003, 15:279-288 349. Greer PA, Meckling-Hansen K, Pawson T: The human c-fps/fes gene product expressed ectopically in rat fibroblasts is nontransforming and has restrained protein-tyrosine kinase activity, Mol Cell Biol 1988, 8:578-587 350. Li J, Smithgall TE: Fibroblast transformation by Fps/Fes tyrosine kinases requires Ras, Rac, and Cdc42 and induces extracellular signal-regulated and c-Jun N-terminal kinase activation, J Biol Chem 1998, 273:13828-13834 351. Cheng HY, Schiavone AP, Smithgall TE: A point mutation in the N-terminal coiled-coil domain releases c-Fes tyrosine kinase activity and survival signaling in myeloid leukemia cells, Mol Cell Biol 2001, 21:6170-6180 352. Paulson R, Jackson J, Immergluck K, Bishop JM: The DFer gene of Drosophila melanogaster encodes two membrane-associated proteins that can both transform vertebrate cells, Oncogene 1997, 14:641-652 353. Kim J, Feldman RA: Activated Fes protein tyrosine kinase induces terminal macrophage differentiation of myeloid progenitors (U937 cells) and activation of the transcription factor PU.1, Mol Cell Biol 2002, 22:1903-1918 354. Kim J, Ogata Y, Feldman RA: Fes tyrosine kinase promotes survival and terminal granulocyte differentiation of factor-dependent myeloid progenitors (32D) and activates lineage-specific transcription factors, J Biol Chem 2003, 278:14978-14984 355. Ferrari S, Manfredini R, Tagliafico E, Grande A, Barbieri D, Balestri R, Pizzanelli M, Zucchini P, Citro G, Zupi G, et al.: Antiapoptotic effect of c-fes protooncogene during granulocytic differentiation, Leukemia 1994, 8 Suppl 1:S91-94 356. Smithgall TE, Yu G, Glazer RI: Identification of the differentiation-associated p93 tyrosine protein kinase of HL-60 leukemia cells as the product of the human c-fes locus and its expression in myelomonocytic cells, J Biol Chem 1988, 263:15050-15055

174

357. Yu G, Smithgall TE, Glazer RI: K562 leukemia cells transfected with the human c-fes gene acquire the ability to undergo myeloid differentiation, J Biol Chem 1989, 264:10276-10281 358. Angst BD, Marcozzi C, Magee AI: The cadherin superfamily, J Cell Sci 2001, 114:625-626 359. McNeill H, Ozawa M, Kemler R, Nelson WJ: Novel function of the cell adhesion molecule uvomorulin as an inducer of cell surface polarity, Cell 1990, 62:309-316 360. Braga V: Epithelial cell shape: cadherins and small GTPases, Exp Cell Res 2000, 261:83-90 361. Kemler R: From cadherins to catenins: cytoplasmic protein interactions and regulation of cell adhesion, Trends Genet 1993, 9:317-321 362. Cowin P, Rowlands TM, Hatsell SJ: Cadherins and catenins in breast cancer, Curr Opin Cell Biol 2005, 17:499-508 363. Sarrio D, Moreno-Bueno G, Sanchez-Estevez C, Banon-Rodriguez I, Hernandez-Cortes G, Hardisson D, Palacios J: Expression of cadherins and catenins correlates with distinct histologic types of ovarian carcinomas, Hum Pathol 2006, 37:1042-1049 364. van Oort IM, Tomita K, van Bokhoven A, Bussemakers MJ, Kiemeney LA, Karthaus HF, Witjes JA, Schalken JA: The prognostic value of E-cadherin and the cadherin-associated molecules alpha-, beta-, gamma-catenin and p120ctn in prostate cancer specific survival: a long-term follow-up study, Prostate 2007, 67:1432-1438 365. Teuliere J, Faraldo MM, Deugnier MA, Shtutman M, Ben-Ze'ev A, Thiery JP, Glukhova MA: Targeted activation of beta-catenin signaling in basal mammary epithelial cells affects mammary development and leads to hyperplasia, Development 2005, 132:267-277 366. Kinch MS, Clark GJ, Der CJ, Burridge K: Tyrosine phosphorylation regulates the adhesions of ras-transformed breast epithelia, J Cell Biol 1995, 130:461-471 367. Kim L, Wong TW: The cytoplasmic tyrosine kinase FER is associated with the catenin-like substrate pp120 and is activated by growth factors, Mol Cell Biol 1995, 15:4553-4561 368. Sloane JP, Ormerod MG: Distribution of epithelial membrane antigen in normal and neoplastic tissues and it value in diagnostic tumor pathology, Cancer 1981, 47:1786-1795 369. Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen D, Gabbiani G: A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation, J Cell Biol 1986, 103:2787-2796 370. Lampugnani MG, Corada M, Andriopoulou P, Esser S, Risau W, Dejana E: Cell confluence regulates tyrosine phosphorylation of adherens junction components in endothelial cells, J Cell Sci 1997, 110 (Pt 17):2065-2077 371. Greenbaum D, Colangelo C, Williams K, Gerstein M: Comparing protein abundance and mRNA expression levels on a genomic scale, Genome Biol 2003, 4:117 372. Manfredini R, Grande A, Tagliafico E, Barbieri D, Zucchini P, Citro G, Zupi G, Franceschi C, Torelli U, Ferrari S: Inhibition of c-fes expression by an antisense oligomer causes apoptosis of HL60 cells induced to granulocytic differentiation, J Exp Med 1993, 178:381-389 373. Stinnakre MG, Vilotte JL, Soulier S, Mercier JC: Creation and phenotypic analysis of alpha-lactalbumin-deficient mice, Proc Natl Acad Sci U S A 1994, 91:6544-6548

175

374. Vadlamudi RK, Wang RA, Talukder AH, Adam L, Johnson R, Kumar R: Evidence of Rab3A expression, regulation of vesicle trafficking, and cellular secretion in response to heregulin in mammary epithelial cells, Mol Cell Biol 2000, 20:9092-9101 375. Spicer AP, Rowse GJ, Lidner TK, Gendler SJ: Delayed mammary tumor progression in Muc-1 null mice, J Biol Chem 1995, 270:30093-30101 376. Kirito K, Nakajima K, Watanabe T, Uchida M, Tanaka M, Ozawa K, Komatsu N: Identification of the human erythropoietin receptor region required for Stat1 and Stat3 activation, Blood 2002, 99:102-110 377. Sangrar W, Gao Y, Zirngibl RA, Scott ML, Greer PA: The fps/fes proto-oncogene regulates hematopoietic lineage output, Exp Hematol 2003, 31:1259-1267 378. Cui Y, Riedlinger G, Miyoshi K, Tang W, Li C, Deng CX, Robinson GW, Hennighausen L: Inactivation of Stat5 in mouse mammary epithelium during pregnancy reveals distinct functions in cell proliferation, survival, and differentiation, Mol Cell Biol 2004, 24:8037-8047 379. Iavnilovitch E, Groner B, Barash I: Overexpression and forced activation of stat5 in mammary gland of transgenic mice promotes cellular proliferation, enhances differentiation, and delays postlactational apoptosis, Mol Cancer Res 2002, 1:32-47 380. Schwertfeger KL, Richert MM, Anderson SM: Mammary gland involution is delayed by activated Akt in transgenic mice, Mol Endocrinol 2001, 15:867-881 381. Moorehead RA, Fata JE, Johnson MB, Khokha R: Inhibition of mammary epithelial apoptosis and sustained phosphorylation of Akt/PKB in MMTV-IGF-II transgenic mice, Cell Death Differ 2001, 8:16-29 382. Kanda S, Kanetake H, Miyata Y: Downregulation of Fes inhibits VEGF-A-induced chemotaxis and capillary-like morphogenesis by cultured endothelial cells, J Cell Mol Med 2007, 11:495-501 383. Izuhara K, Feldman RA, Greer P, Harada N: Interleukin-4 induces association of the c-fes proto-oncogene product with phosphatidylinositol-3 kinase, Blood 1996, 88:3910-3918 384. Kanda S, Mochizuki Y, Miyata Y, Kanetake H: The role of c-Fes in vascular endothelial growth factor-A-mediated signaling by endothelial cells, Biochem Biophys Res Commun 2003, 306:1056-1063 385. Kanda S, Miyata Y, Mochizuki Y, Matsuyama T, Kanetake H: Angiopoietin 1 is mitogenic for cultured endothelial cells, Cancer Res 2005, 65:6820-6827 386. Mochizuki Y, Nakamura T, Kanetake H, Kanda S: Angiopoietin 2 stimulates migration and tube-like structure formation of murine brain capillary endothelial cells through c-Fes and c-Fyn, J Cell Sci 2002, 115:175-183 387. Arregui C, Pathre P, Lilien J, Balsamo J: The nonreceptor tyrosine kinase fer mediates cross-talk between N-cadherin and beta1-integrins, J Cell Biol 2000, 149:1263-1274 388. El Sayegh TY, Arora PD, Fan L, Laschinger CA, Greer PA, McCulloch CA, Kapus A: Phosphorylation of N-cadherin-associated cortactin by Fer kinase regulates N-cadherin mobility and intercellular adhesion strength, Mol Biol Cell 2005, 16:5514-5527 389. Baki L, Marambaud P, Efthimiopoulos S, Georgakopoulos A, Wen P, Cui W, Shioi J, Koo E, Ozawa M, Friedrich VL, Jr., Robakis NK: Presenilin-1 binds cytoplasmic epithelial cadherin, inhibits cadherin/p120 association, and regulates stability and function of the cadherin/catenin adhesion complex, Proc Natl Acad Sci U S A 2001, 98:2381-2386

176

390. Wildenberg GA, Dohn MR, Carnahan RH, Davis MA, Lobdell NA, Settleman J, Reynolds AB: p120-catenin and p190RhoGAP regulate cell-cell adhesion by coordinating antagonism between Rac and Rho, Cell 2006, 127:1027-1039 391. Drayton S, Peters G: Immortalisation and transformation revisited, Curr Opin Genet Dev 2002, 12:98-104 392. Dimri G, Band H, Band V: Mammary epithelial cell transformation: insights from cell culture and mouse models, Breast Cancer Res 2005, 7:171-179 393. Heinrichs S, Deppert W: Apoptosis or growth arrest: modulation of the cellular response to p53 by proliferative signals, Oncogene 2003, 22:555-571 394. Sionov RV, Haupt Y: The cellular response to p53: the decision between life and death, Oncogene 1999, 18:6145-6157 395. Bourdon JC: p53 Family isoforms, Curr Pharm Biotechnol 2007, 8:332-336 396. Vogelstein B, Lane D, Levine AJ: Surfing the p53 network, Nature 2000, 408:307-310 397. Kiyokawa H: Senescence and cell cycle control, Results Probl Cell Differ 2006, 42:257-270 398. Hanahan D, Weinberg RA: The hallmarks of cancer, Cell 2000, 100:57-70 399. Hoogervorst EM, van Steeg H, de Vries A: Nucleotide excision repair- and p53-deficient mouse models in cancer research, Mutat Res 2005, 574:3-21 400. Kimbro KS, Simons JW: Hypoxia-inducible factor-1 in human breast and prostate cancer, Endocr Relat Cancer 2006, 13:739-749 401. Lavin MF, Gueven N: The complexity of p53 stabilization and activation, Cell Death Differ 2006, 13:941-950 402. Brooks CL, Gu W: Ubiquitination, phosphorylation and acetylation: the molecular basis for p53 regulation, Curr Opin Cell Biol 2003, 15:164-171 403. Appella E, Anderson CW: Post-translational modifications and activation of p53 by genotoxic stresses, Eur J Biochem 2001, 268:2764-2772 404. Ito A, Lai CH, Zhao X, Saito S, Hamilton MH, Appella E, Yao TP: p300/CBP-mediated p53 acetylation is commonly induced by p53-activating agents and inhibited by MDM2, Embo J 2001, 20:1331-1340 405. Shaw PH: The role of p53 in cell cycle regulation, Pathol Res Pract 1996, 192:669-675 406. Shu KX, Li B, Wu LX: The p53 network: p53 and its downstream genes, Colloids Surf B Biointerfaces 2007, 55:10-18 407. Schuler M, Green DR: Mechanisms of p53-dependent apoptosis, Biochem Soc Trans 2001, 29:684-688 408. Su JD, Mayo LD, Donner DB, Durden DL: PTEN and phosphatidylinositol 3'-kinase inhibitors up-regulate p53 and block tumor-induced angiogenesis: evidence for an effect on the tumor and endothelial compartment, Cancer Res 2003, 63:3585-3592 409. Toschi E, Rota R, Antonini A, Melillo G, Capogrossi MC: Wild-type p53 gene transfer inhibits invasion and reduces matrix metalloproteinase-2 levels in p53-mutated human melanoma cells, J Invest Dermatol 2000, 114:1188-1194 410. Sherr CJ, DePinho RA: Cellular senescence: mitotic clock or culture shock? Cell 2000, 102:407-410 411. Roemer K: Mutant p53: gain-of-function oncoproteins and wild-type p53 inactivators, Biol Chem 1999, 380:879-887 412. vom Brocke J, Schmeiser HH, Reinbold M, Hollstein M: MEF immortalization to investigate the ins and outs of mutagenesis, Carcinogenesis 2006, 27:2141-2147

177

413. McCaffrey AP, Meuse L, Pham TT, Conklin DS, Hannon GJ, Kay MA: RNA interference in adult mice, Nature 2002, 418:38-39 414. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC: Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans, Nature 1998, 391:806-811 415. Dirac AM, Bernards R: Reversal of senescence in mouse fibroblasts through lentiviral suppression of p53, J Biol Chem 2003, 278:11731-11734 416. Wiznerowicz M, Szulc J, Trono D: Tuning silence: conditional systems for RNA interference, Nat Methods 2006, 3:682-688 417. Jerry DJ, Medina D, Butel JS: p53 mutations in COMMA-D cells, In Vitro Cell Dev Biol Anim 1994, 30A:87-89 418. Kumar S, Walia V, Ray M, Elble RC: p53 in breast cancer: mutation and countermeasures, Front Biosci 2007, 12:4168-4178 419. Dumble ML, Croager EJ, Yeoh GC, Quail EA: Generation and characterization of p53 null transformed hepatic progenitor cells: oval cells give rise to hepatocellular carcinoma, Carcinogenesis 2002, 23:435-445 420. Harvey M, Sands AT, Weiss RS, Hegi ME, Wiseman RW, Pantazis P, Giovanella BC, Tainsky MA, Bradley A, Donehower LA: In vitro growth characteristics of embryo fibroblasts isolated from p53-deficient mice, Oncogene 1993, 8:2457-2467 421. Tsukada T, Tomooka Y, Takai S, Ueda Y, Nishikawa S, Yagi T, Tokunaga T, Takeda N, Suda Y, Abe S, et al.: Enhanced proliferative potential in culture of cells from p53-deficient mice, Oncogene 1993, 8:3313-3322 422. Vodicka MA: Determinants for lentiviral infection of non-dividing cells, Somat Cell Mol Genet 2001, 26:35-49 423. Agiostratidou G, Hulit J, Phillips GR, Hazan RB: Differential cadherin expression: potential markers for epithelial to mesenchymal transformation during tumor progression, J Mammary Gland Biol Neoplasia 2007, 12:127-133 424. Ivaska J, Pallari HM, Nevo J, Eriksson JE: Novel functions of vimentin in cell adhesion, migration, and signaling, Exp Cell Res 2007, 313:2050-2062 425. McInerney JM, Nawrocki JR, Lowrey CH: Long-term silencing of retroviral vectors is resistant to reversal by trichostatin A and 5-azacytidine, Gene Ther 2000, 7:653-663 426. Sarrio D, Rodriguez-Pinilla SM, Hardisson D, Cano A, Moreno-Bueno G, Palacios J: Epithelial-mesenchymal transition in breast cancer relates to the basal-like phenotype, Cancer Res 2008, 68:989-997 427. Botelho RJ, Teruel M, Dierckman R, Anderson R, Wells A, York JD, Meyer T, Grinstein S: Localized biphasic changes in phosphatidylinositol-4,5-bisphosphate at sites of phagocytosis, J Cell Biol 2000, 151:1353-1368 428. Mitsui N, Inatome R, Takahashi S, Goshima Y, Yamamura H, Yanagi S: Involvement of Fes/Fps tyrosine kinase in semaphorin3A signaling, Embo J 2002, 21:3274-3285 429. Yu G, Grant S, Glazer RI: Association of p93c-fes tyrosine protein kinase with granulocytic/monocytic differentiation and resistance to differentiating agents in HL-60 leukemia cells, Mol Pharmacol 1988, 33:384-388 430. Tu ZJ, Kollander R, Kiang DT: Differential up-regulation of gap junction connexin 26 gene in mammary and uterine tissues: the role of Sp transcription factors, Mol Endocrinol 1998, 12:1931-1938

178

431. Tu ZJ, Pan W, Gong Z, Kiang DT: Involving AP-2 transcription factor in connexin 26 up-regulation during pregnancy and lactation, Mol Reprod Dev 2001, 59:17-24 432. Zhang J, Brewer S, Huang J, Williams T: Overexpression of transcription factor AP-2alpha suppresses mammary gland growth and morphogenesis, Dev Biol 2003, 256:127-145 433. Johansson EM, Kannius-Janson M, Bjursell G, Nilsson J: The p53 tumor suppressor gene is regulated in vivo by nuclear factor 1-C2 in the mouse mammary gland during pregnancy, Oncogene 2003, 22:6061-6070 434. De Matteis MA, Di Campli A, Godi A: The role of the phosphoinositides at the Golgi complex, Biochim Biophys Acta 2005, 1744:396-405 435. Weixel KM, Blumental-Perry A, Watkins SC, Aridor M, Weisz OA: Distinct Golgi populations of phosphatidylinositol 4-phosphate regulated by phosphatidylinositol 4-kinases, J Biol Chem 2005, 280:10501-10508 436. Bard F, Malhotra V: The formation of TGN-to-plasma-membrane transport carriers, Annu Rev Cell Dev Biol 2006, 22:439-455 437. McMahon HT, Gallop JL: Membrane curvature and mechanisms of dynamic cell membrane remodelling, Nature 2005, 438:590-596 438. Zimmerberg J, Kozlov MM: How proteins produce cellular membrane curvature, Nat Rev Mol Cell Biol 2006, 7:9-19 439. Heid HW, Keenan TW: Intracellular origin and secretion of milk fat globules, Eur J Cell Biol 2005, 84:245-258 440. Desai A, Mitchison TJ: Microtubule polymerization dynamics, Annu Rev Cell Dev Biol 1997, 13:83-117 441. Shelden E, Wadsworth P: Observation and quantification of individual microtubule behavior in vivo: microtubule dynamics are cell-type specific, J Cell Biol 1993, 120:935-945 442. Surrey T, Nedelec F, Leibler S, Karsenti E: Physical properties determining self-organization of motors and microtubules, Science 2001, 292:1167-1171 443. Pepperkok R, Bre MH, Davoust J, Kreis TE: Microtubules are stabilized in confluent epithelial cells but not in fibroblasts, J Cell Biol 1990, 111:3003-3012 444. Kogata N, Masuda M, Kamioka Y, Yamagishi A, Endo A, Okada M, Mochizuki N: Identification of Fer tyrosine kinase localized on microtubules as a platelet endothelial cell adhesion molecule-1 phosphorylating kinase in vascular endothelial cells, Mol Biol Cell 2003, 14:3553-3564 445. Hartsock A, Nelson WJ: Adherens and tight junctions: Structure, function and connections to the actin cytoskeleton, Biochim Biophys Acta 2008, 1778:660-669 446. Medina D, Kittrell FS: Establishment of mouse mammary cell lines. In: M.M. Ip and B.B. Ash, Editors, Methods in Mammary Gland Biology and Breast Cancer, Kluwer Academic/Plenum Publishers, New York (2000), pp. 137-145 (Chapter 13).