Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Role of the cytoskeleton in basement membrane-induced mammary epithelial morphogenesis and differentiation Somasiri, Aruna Mahendra 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-0615.pdf [ 7.74MB ]
Metadata
JSON: 831-1.0089197.json
JSON-LD: 831-1.0089197-ld.json
RDF/XML (Pretty): 831-1.0089197-rdf.xml
RDF/JSON: 831-1.0089197-rdf.json
Turtle: 831-1.0089197-turtle.txt
N-Triples: 831-1.0089197-rdf-ntriples.txt
Original Record: 831-1.0089197-source.json
Full Text
831-1.0089197-fulltext.txt
Citation
831-1.0089197.ris

Full Text

ROLE OF THE CYTOSKELETON IN BASEMENT MEMBRANE-INDUCED MAMMARY EPITHELIAL MORPHOGENESIS AND DIFFERENTIATION by ARUNA MAHENDRA SOMASIRI B.Sc. (Hons. Biochem), The University of British Columbia, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTERS OF SCIENCE in FACULTY OF GRADUATE STUDIES (DEPARTMENT OF ANATOMY) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 1999 © Aruna Mahendra Somasiri UBC Special Collections - Thesis Authorisation Form http://www.Ubrary.ubc xa/spcoMhesauth.html In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the re q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g of ' t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the head of my department o r by h i s o r her r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department of The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Date 0(\ 17 \S31 1 of 1 10/12/99 2:43 PM A B S T R A C T Basement membrane-mediated mammary epithelial morphogenesis is characterized by both morphological changes to the cells and the induction of milk protein gene expression. When mammary epithelial cells are placed on a reconstituted basement membrane gel, they aggregate, pull the gel around them and cavitation takes place. During this morphogenic process, the cells undergo changes in shape, polarity and cell-cell junctions. In this thesis I have sought to identify the role of the three major cytoskeletal elements in theses changes and to determine their involvement in the associated induction of milk protein gene induction. When scp2 cells were placed on Matrigel, they formed polarized structures with a central lumen that resemble functional mammary alveolar in vivo. Within these "mammospheres" E-cadherin was localized to adherens junctions, occluding was localized to tight junctions and actin filaments formed an apical junction-associated network. These mammospheres differentiated and expressed two milk proteins: lactoferrin, which is transcriptionally regulated by cell rounding only and P-casein, which is transcriptionally regulated by a6p4 integrin-mediated morphogenic changes initiated by the ligation to laminin within the Matrigel. The cells cultured on Matrigel were treated with increasing concentrations of nocodazole (ND), acrylamide (Ac) and cytochalasin D (CD) for 4 hours and allowed to under go morphogenesis for 72 hours. All three disrupting agents prevented the cells from undergoing morphogenesis and inhibited differentiation. ND treatment inhibited the synthesis of both lactoferrin and fi-casein while Ac and CD treatments inhibited the p-ii casein induction only. These observations suggested that ND was not specifically inhibiting differentiation by preventing morphogenesis. In contrast, Ac and CD appeared to inhibit differentiation in a morphogenesis-dependent manner. The 0:604 integrin becomes physically linked to keratin intermediate filaments after its ligation to laminin. Because cc6p4 ligation also initiates apical/basal polarity, cells on Matrigel were treated with a short term high-dose of CD to completely disrupt f-actin and were then observed for a six day recovery period. The apical actin network and tight junctions were re-established first followed by the induction of P-casein. Therefore, keratin and actin appear to influence functional mammosphere formation by helping to initiate and maintain cell polarity. iii TABLE OF CONTENTS ABSTRACT ii LIST O F TABLES viii LIST OF FIGURES ix LIST OF ABBREVIATIONS xi ACKNOWLEDGMENTS xii C H A P T E R I. I N T R O D U C T I O N 1 1.1. Mouse mammary epithelial morphogenesis and differentiation in vivo 2 1.2. Mouse mammary epithelial morphogenesis and differentiation in culture 5 1.3. Milk protein gene expression 8 1.4. Basement membrane (BM) 9 1.5. Intergrins 10 1.6. The cytoskeleton 11 a. Microfilaments (actin) 11 b. Intermediate filaments (IF) 13 c. Microtubules 14 1.7. Cell-Cell and Cell-ECM Junctions 15 a. Tight Junctions. 16 b. Adherens Junctions 18 c. Desmosomes 18 d. Hemidesmosome 19 1.8. Hypothesis 25 C H A P T E R H . M A T E R I A L S A N D M E T H O D S 28 1. Cell line 28 iv 2. Cell culture 2.1. Maintenance of a pure mammary epithelial cell population 2.2. MATRIGEL® Basement Membrane Matrix (EHS) 2.3. Preparation of MATRIGEL coated dishes 2.4. Differentiation assay 3. Cytoskeletal disruption assays 3.1. Monolayer cultures a. Microtubule disruption assay b. Keratin disruption assay c. Actin disruption assay 3.2. Cells cultured on Matrigel a. Microtubule disruption assay b. Keratin disruption assay c. Actin disruptions 4. Western blot analysis 4.1. For fj-casein and lactoferrin 4.2. For tight junction protein occluding 5. Immunofluorescence Microscopy 5. l.Tubulin staining 5.2. Keratin staining 5.3. Actin staining 5.4. E-cadherin staining 5.5. Occludin staining 5.6. Image collection CHAPTER m. RESULTS 1. Optimal conditions to disrupt cytoskeletal components: microtubules, keratin and actin 1.1. Microtubule disruption 1.2. Intermediate filament disruption 1.3. Actin filament disruption 42 2. Disruption specificity 49 2.1. Nocodzole treatment 49 2.2. Acrylamide treatment. 49 2.3. Cytochalasin D treatment 50 3. Normal morphogenesis and differentiation of mouse mammary epithelialcells placed on a basement membrane 52 3.1 Morphogenesis 52 3.2 Differentiation 53 4. Effects of nocodazol treatment on the morphogenesis and differentiation of cells cultured on Matrigel 58 4.1 Effects on cell morphogenesis 58 4.2. Effects on cell differentiation 59 5. Effects of acrylamide treatment on the morphogenesis and differentiation of cells cultured on Matrigel 64 5.1. Effects on cell morphogenesis and keratin filaments 64 5.2. Effects on cell differentiation 65 6. Effects of cytochalasin D treatment on the morphogenesis and differentiation of cells cultured on Matrigel 70 6.1. Effects on cell morphogenesis 70 6.2. Effects on cell differentiation 70 7. Effects of cytochalasin D treatment on morphogenesis-dependent cell-cell junction formation 78 7.1 Cell-cell junctions of cell monolayers and mammosphers 78 7.2 The effect of cytochalasin treatment on tight junction formation 79 7.3 Cell shape regulates tight junction formation 80 CHAPTER IV. DISCUSSION 92 1. Mammosphere formation and differentiation on reconstituted basement membrane (Matrigel) 92 2. Effects of nocodazole 92 vi 3. Effects of acrylamide 94 4. Effects of cytochalasin D 96 5. Tight junction formation is correlated with P-casein induction 97 6. Conclusions. 99 V. R E F E R E N C E S 102 vii L I S T O F T A B L E S Table # Table 1. Table 2. Page # Optimal doses, times and effects of the disrupting agents on each individual cytoskeletal element and cell morphology 48 Effects of each disrupting agent on the microtubule, keratin And actin cytoskeletons 51 viii L I S T O F F I G U R E S Figure* Page# Fig. 1 Junctional complexes in epithelial cells 22 Fig. 2 Integrin-intermediate filament interactions (Hemidesmosomes) 24 Fig. 3 Mouse mammary epithelial cell in culture 27 Fig. 4 Scp2 mammary epithelial cell monolayers treated with disrupting Agents 45 Fig. 5 Morphological effects of cytoskeletal disruption on cell monolayers 47 Fig. 6 Basement membrane-dependent mouse mammary epithelial cell Morphogenesis 55 Fig. 7 Mouse mammary epithelial cell differentiation on Matrigel 57 Fig. 8 Cell cluster morphology of cells treated with nocodazole and recovered for 72 hours 61 Fig. 9 Lactoferrin and (3-casein expression in cells cultured on Matrigel and treated with nocodazole 63 Fig. lOCell cluster morphology of cells treated with acrylamide and recovered for 72 hours 67 Fig. 11 Lactoferrin and fi-casein expression in cells plated on Matrigel and treated with acrylamide 69 Fig. 12 Cell cluster morphology of cells treated with cytochalasin D and recovered for 72 hours 73 Fig. 13 Actin localization of cell clusters treated with cytochalasin D and recovered for 72 hr 75 Fig. 14 Lactoferrin and (3-casein expression in cells cultured on Matrigel treated with cytochalasin D 77 Fig. 15 Actin, E-cadherin and occludin localization in mammary epithelial cell monolayers and mammospheres 83 ix Fig. 16 Actin ctyosleleton in cells plated on Matrigel, treated with lfJLig/ml cytochalasin D and allowed to recover for 6 days Fig. 170ccludin in cells plated on Matrigel, treated with 10p,g/ml cytochalasin D and allowed to recover for 6 days Fig. 18 Lactoferrin and (3-casein expression in cells plated on Matrigel, treated with CD and recovered for 7 days Fig. 19 Occludin phosphorylation and membrane localization Fig. 20 Summary model for possible role of the cytoskeleton in mammary epithelial morphogenesis and differentiation L I S T O F A B B R E V I A T I O N S Ac Acrylamide APC Adenomatous polyposis B M Basement Membrane BPAG Bullous pemphigoid antigen BSA , Bovine serum albumin CD Cytochalasin D DAPI 4'-6-diamidine-2-phenyl indole DC Desmocolin DL Desmoglein DMEM Dulbecco's modified Eagle's medium DMSO Dimethyl sulfoxide DOC Sodium deoxycholate ECL Enhanced chemiluminescent ECM Extracellular Matrix EHS Engelberth-Holm-Swarm FBS Fetal bovine serum Fn Fibronectin HBSS Hanks balanced salt solution HGF Hepatocyte growth factor IF Intermediate filaments LTF Lactoferrin NGS Normal goat serum ND Nocodazole NDF neu differentiation factor PAGE Polyacrylamide gel electrophoresis PBS Phosphate buffered saline PMSF Phenylmethyl-sulfonyl fluoride SDS Sodium dodecyl sulphate TBS Tris-buffered saline TGF-p Transforming growth factor TJ Tight junctions x i A C K N O W L E D G M E N T S I am happy to have this opportunity to thank the people who have helped and guided me in numerous ways. Dr Calvin Roskelley, my supervisor, for his perseverance in teaching me the principles of cell biology, his enthusiasm and encouragement throughout my times in his laboratory and for labouring through the numerous drafts of this thesis. It has been a great privilege to work with him and I look forward to continuing with related research in his lab. I would also like to thank the other members of my committee, Drs. Wayne Vogl, Donald Moerman and Fumio Takei for both their help and guidance throughout this presented work. I would also like to thank Krista McCutcheon for introducing me to the department of Anatomy and to the lab. The faculty & staff and the fellow graduate students of the department also deserve a thank for making my stay in the department a memorable one. A special thank for Colleen Wu and Jie Pan for all of their help in the lab and Arthur Legg for proof reading numerous drafts of this thesis. Finally, I would like to especially thank my mom, dad and brother, without whose support and tolerance I would never have been able to complete or begin my studies in Canada. Sorry, for not being able to visit you more often as you like me to. This work was supported by a grant to Dr. C. D. Roskelley from the National Cancer Institute of Canada. xii I. INTRODUCTION The mammary gland represents an excellent biological system to study the, mechanisms which regulate epithelial cell development. Unlike most organs, where development takes place during embryogenesis and fetal stages, the mammary gland is capable of undergoing developmental cycles throughout adult reproductive life (Daniel and Silberstein, 1987). Each cycle can be described in the following manner: early in pregnancy, terminal mammary epithelial endbuds undergo a massive proliferative burst which is followed by a controlled branching; later in pregnancy alveolar epithelium differentiate, begin to express milk protein genes and alveolar structures form; during lactation milk proteins and associated carbohydrates and lipids are vectorially secreted into the central alveolar lumen; during involution, milk protein gene expression ceases, the alveolar epithelium becomes disorganized and apoptosis occurs which ceases much of the alveolar epithelium and returns the gland to its original state. From this brief description it is clear that morphogenetic changes (ie. formation of the mammary alveolai) are tightly coupled to differentiative changes (ie. tissue specific gene expression) throughout the cycle. Therefore, I hypothesize that both morphogenesis and differentiation are regulated, at least in part, by the cytoskeleton. To test this hypothesis I have used three-dimensional culture system to examine the formation of alveoli and the associated induction of differentiative milk protein expression. 1 1.1. Mouse mammary epithelial morphogenesis and differentiation in vivo. At birth, the parenchyma of the mouse mammary gland consists of branching cords of ectodermally derived epithelial cells connected to the nipple through a single primary duct (Sakakura, 1987). Prior to puberty, these epithelial cords cavitate to form ducts with a central lumen. Throughout their length, the ducts are surrounded by a sleeve of fibroblastic stroma except at their terminal ends. The latter structures are known as end buds and they directly contact the gland's highly adipocytic stroma; Interactions between the epithelium and the mesenchymally-derived stroma are the major determinants of tissue-specific mammary gland morphogenesis. When mammary epithelium is combined with the salivary gland mesenchyme, it develops salivary gland ductal patterns (Kratochwil, 1969). Co-cultures of mammary epithelial cells and stromal fibroblasts cause extracellular matrix (ECM) deposition (Reichman et al., 1986). Therefore, it is possible that the ECM deposition induced by the stromal fibroblasts may regulate epithelial cell fate. Prior to puberty, the rate of mammary gland growth parallels the growth of the rest of the body. During puberty, however, endbuds and connective tissue increase their growth rates under the influence of estrogen produced by the ovaries (Daniel and Silberstein., 1987). Increased cell proliferation at the end buds leads to the elongation of the ductal system. At the onset of pregnancy, as levels of estrogen and progesterone increase, there is massive cell proliferation in the end buds, which leads to the displacement of much of the surrounding adipocytic stroma to the margins of the gland (Topper and Freeman, 1980). 2 Lateral branching of the end buds is responsible for the tree like-pattern characteristic of the parenchyma of the adult mammary gland. Dominant negative mutants of transforming growth factor-beta (TGF-p; Joseph et al., 1999) or TGF-(3 receptor (Gorska et al., 1998) in transgenic mice leads to increased lateral branching and hyperplasia of the end buds in the virgin mammary ductal system. Thus, TGF-fi functions as a localized epithelial growth inhibitor to prevent chronic lateral budding and to maintain the normal pattern of branching during ductal morphogenesis. TGF-fJ expression is regulated at the transcriptional level by the specialized basement membrane extracellular matrix (ECM) that surrounds the mammary epithelium (Streuli et al, 1993). In transgenic mice in which the expression of an autoactivated form of the entactin degrading metalloproteinase stromelysin-1 is targeted to the mammary epithelium, the basement membrane is constantly remodeled and end bud hyperbranching occurs (Sympson et al., 1994; Sternlicht et al., 1999). Therefore, an intact basement membrane is a critical regulator of early mammary-specific morphogenetic events. Early in the pregnancy, the proliferating cells of endbuds undergo a secondary, spatially restricted branching that is regulated by hepatocyte growth factor (HGF; Neimann et al., 1998). HGF is another locally acting growth factor that is produced by mammary mesenchymal cells, is sequestered in the ECM and acts by binding to the c-met receptor tyrosine kinase on the surface of epithelial cells. In organ cultures, treatment with antisense oligonucleotides against HGF completely abolishes the branching morphology (Yang et al., 1995). Conversely, overexpression of HGF and or the c-Met receptor in transgenic animals leads to enhanced end bud size and number, and it induces hyperplastic branching morphology (Yant et al., 1998). 3 Later in pregnancy endbuds expand to form alveoli which is under the control of the neu differentiation factor (NDF; Neimann et al., 1998). NDF is another morphogen which is produced by the mammary mesenchyme, is sequestered in the ECM and acts on epithelial receptors, in this case the erbB family of tyrosine kinases. In mammary organ cultures derived from pregnant mice NDF antisense oligonucleotides inhibit alveolar morphogenesis, a phenotype that can be rescued by the addition of recombinant NDF. In addition, in virgin transgenic mice overexpressing mammary-targetted NDF, endbuds undergo premature endbud expansion and alveolar morphogenesis (Krane et al., 1996). Importantly, in all cases, when alveolar morphogenesis is affected epithelial differentiation is similarly affected. Thus, when NDF-mediated morphogenesis is blocked, differentiation does not occur. Mammary epithelial differentiation is marked by the induction of milk protein gene expression. This differentiation is regulated by NDF, lactogenic hormones and the basement membrane surrounding the developing alveolus (Simpson et al., 1998; Howlett and Bissell, 1993; Teng et al., 1989; Lee et al., 1998). The major mouse milk proteins are: the ion-binding proteins, one of which is lactoferrin, that are first expressed in early to mid pregnancy; the caseins, one of which is P-casein, that are first expressed in mid to late pregnancy; and the whey proteins, one of which is whey acidic protein (WAP) which is first expressed in late pregnancy. Because I am interested in the morphogenetic formation of alveoli, which is essentially complete by mid-pregnancy, I have focused on the regulation of lactoferrin and P-casein. 4 1.2. Mouse mammary epithelial morphogenesis and differentiation in culture When differentiated, mammary alveolar epithelial cells from mid pregnant mice are enzymatically dissociated and cultured on tissue culture plastic, they loose their alveolar morphology and dedifferentiate to form a confluent monolayer of polygonal shaped'cells (Emerman and Pitelka, 1977). These epithelial cells display no intracellular polarity, have a poorly developed secretory apparatus, and they have a high nuclear to cytoplasmic ratio. In addition, these monolayers are treated with lactogenic hormones insulin, Cortisol and prolactin, no milk protein gene induction is observed (Emermann et al., 1977; Lee etal., 1984). When these epithelial cells are placed on attached collagen gels they form thin continuous monolayers that are indistinguishable from the monolayers on plastic dishes. However, if the collagen gels are made flexible by floating them in the media they contract and the cells begin to round-up (Emerman and Pitelka, 1977; Emerman et al., 1979). With time, the cells on the floating gels become columnar in shape and their nuclei become basally located. At the ultrastructural level, the rough endoplasmic reticulum occupies a larger cytoplasmic volume when compared to cells grown on either plastic or on attached gels, and apical microvilli as well as cell-cell junctional complexes form (Emerman and Pitelka, 1977). When these cells on floating collagen gels are treated with lactogenic hormones insulin, Cortisol and prolactin, there is a significant induction of milk protein gene expression. Therefore, morphologic and functional differentiation are regulated by the flexibility of the collagen gel (Lee et al., 1984). 5 Primary mammary epithelial cells plated on either floating collagen gels or on tissue culture plastic produce significant amounts of ECM proteins. However, only the cells on floating gels are able to correctly polarize and deposit these proteins appropriately in a basal location (Streuli and Bisseil, 1990). In fact, cells on the floating gel deposit an intact basement membrane which induces heparan sulfate (Parry et al., 1982), type IV collagen and laminin (Parry et al., 1985; Streuli and Bisseil, 1990). In vivo, the differentiating mammary epithelium is in contact with a continuous basement membrane (Silberstein and Daniel, 1982; Warburton et al., 1982, 1984). To determine if basement membrane deposition observed in floating collagen gel cultures is responsible for mammary epithelial cell morphogenesis and differentiation, these cells have been placed directly on reconstituted 'Engelbreth-Holm-Swarm' tumor-derived matrices (Matrigel; Kleinman et al., 1986). Matrigel contains laminin, type IV collagen, heparan sulphate, protoglyans and entactin. When mammary epithelial cells are initially placed on Matrigel, they form small aggregates. With time, the aggregates "pull" the flexible EHS matrix around them. After two days in culture, these Matrigel covered aggregates cavitate and by day 4, cells become apically/basally polarized as seen in the alveoli of a midpregnant or lactating mammary gland (Barcellos-Hoff et al, 1989; Aggeler et al., 1991). Thus, like alveoli in vivo, these three-dimensional spherical structures in culture contain morphologically polarized cells that rest on a basement membrane with the apical surface facing the central lumen. Therefore, we have called these structures 'mammospheres' (Somasiri and Roskelley, 1999). Ultrastructural analysis of these mammospheres indicates that each individual cell facing the central lumen is connected by apical junctional complexes and EGTA treatment studies 6 demonstrate that these complexes contain tight junctions (Barcellos-Hoff et al., 1989; Pitelka et al., 1983). In addition, adherens junctions are also important in the formation of these alveolar-looking structures (Lochter et al., 1997). In the presence of lactogenic hormones, mammospheres express milk proteins, form secretory vesicles containing micelles, and vectorially release casein containing vesicles and lipid droplets in to the central lumen (Aggeler et al., 1991). Therefore Matrigel matrix provides both the mechanical flexibility for marphogenesis and cell-ECM interactions for the specific signals that initiate mammary epithelial cell morphogenesis and differentiation. As discussed above, the process of mammary epithelial morphogenesis requires alterations in cell shape, cell-cell interactions, and cell-ECM interactions, all of which are important cellular processes influenced by the cytoskeleton. Cell shape is most often influenced by changes to the actin-based microfilament cytoskeleton (Ingber et al., 1995). Cell-cell interactions, particularly those that are mediated by adherens junctions and tight junctions, are also heavily influenced by actin (Citi et al., 1994). The same is true for most cell-ECM interactions that are mediated by integrins, one exception being the oc6(34 integrin which forms a physical link between the basement membrane and the keratin intermediate filament cytoskeleton (Dowling et al., 1996). While microtubules have not been directly implicated in any of these cellular processes, they are required for polarity-based vectorial secretion. Thus, microtubule disruption prevents milk protein accumulation in mammary alveolar central lurnina (Kundsen et al., 1978). A major goal of this thesis is to determine the role of each of these three major cytoskeletal elements in specific aspects of mammary morphogenesis and to determine if they also influence differentiative milk protein gene induction. 7 1.3. Milk protein gene expression The ability of cultured mammary epithelial cells to express rnilk proteins varies depending on the nature of the ECM and the level of morphogenesis attained. Cell monolayers maintained on tissue culture plastic or on attached collagen gels do not express milk proteins except for very low levels of transferrin. When the collagen gels are made flexible by floating them in the medium, the cells express transferrin, lactoferrin and caseins (Lee et al., 1984). Thus, lactoferrin and casein induction are correlated with basement membrane deposition, cell rounding and the initiation of cell polarity. When they are on Matrigel, over 90% of the cells express transferrin, lactoferrin and caseins (Li et al., 1987). In addition, when the mammospheres are fully formed, the cells begin to express the whey proteins (Lin et al., 1995). Thus, whey protein induction is correlated with the secretory phenotype. In this project, I have focused exclusively on morphogenesis rather than secretion. Therefore, I specifically examined the induction of lactoferrin and P-casein. Lactoferrin (LTF) is a 71 kDa iron binding milk protein. Unlike P-caseins lactoferrin is expressed in both non-pregnant and pregnant animals, and its induction does not require the lactogenic hormone prolactin (Teng et al, 1989; Close et al, 1997). The mouse mammary epithelial cell clone scp2 when grown as monolayers on tissue culture plastic, do not express lactoferrin. When these cells are placed on Matrigel or rounded up by placing on dishes coated with non-adhesive substratum polyHEMA, in the absence of exogenously added ECM, lactoferrin expression is induced. Therefore, lactoferrin 8 induction is regulated by changes in the cell shape. Furthermore, this induction is regulated transcriptionally and a 2.6 Kb fragment of the lactoferrin promoter is activated by cell rounding (Close et al., 1997). Thus, lactoferrin protein accumulation accurately reflects the induction state of the gene (Close et al., 1997). P-Casein is a 30 kDa phosphoprotein that is first expressed in mid-pregnant animals. As is the case for lactoferrin, cell rounding is required for P-casein induction, but it is not sufficient (Roskelley et al., 1994). In addition, the lactogenic hormone prolactin is required as in the specific interaction between oc6P4 integrin and laminin (Muschler et al., 1999). Due to the later interaction, in three-dimensional cultures, P-casein induction only occurs when morphogenesis is initiated by Matrigel (Somasiri and Roskelley, 1999). This induction is transcriptionally regulated and a 160 bp element in the P-casein promoter has been identified that is both prolactin and basement membrane-responsive (Schmidhauser et al., 1990; Myers et al, 1998). Thus, protein accumulation accurately reflects the induction state of the P-casein gene (Muschler et al., 1999) 1.4. Basement membrane (BM) Basement membrane (BM) or basal lamina is an organized complex form of the ECM between epithelial cells and the surrounding stromal compartment. The ECM proteins in the basement membrane include laminin, collagen, proteoglycans and glycoproteins (Silberstein and Daniel, 1982; Warburton et al., 1982, 1984). All of these molecules can independently interact with cell surface receptor integrins. These interactions not only account for morphological effects on cells, but are also important in providing signaling cues that regulate growth, differentiation and gene expression. The 9 B M can be divided in to three major zones: 1.) the lamina lucida externa, which an electron lucent region 20-40 nm wide and found just bellow the epithelial basal cell surface, 2.) the lamina densa, a middle layer 20-100 nm wide containing a filamentous meshwork that gives it an electron dense appearance, 3.) the lamina lucida interna, which is an electron lucent region of variable thickness and is found between the lamina densa and the underlying connective tissue. (Martin et al., 1982) Laminin, is located primarily in the lamina densa which becomes periodically fused to the basal surface of the epithelial cells at hemidesmosomal "spotwelds" consists of one heavy a chain (400 kDa) and two light (3 and y chains (130-200 kDa). There are five different a chains, three (3 chains and two y chains, which organize in the form of a cross to make eleven different known kminin molecules. Laminin variants are expressed in a tissue specific and a developmentally regulated manner (Paulsson et al.., 1991). The mammary epithelial basement membrane contains laminin 1, 4 and 5 while Matrigel contains only laminin 1. 1.5. Integrins Integrins are a family of cell surface receptors that mediate adhesive interactions with other cells and with extracellular matrix proteins. These interactions are involved in the regulation of many cellular functions, including development, differentiation, tumor cell growth and metastasis, apoptosis, cell migration and gene activation (Dedhar and Hannigan, 1996; Gumbiner, 1996; Hynes 1992). These trans-membrane glycoproteins are composed of non-covalently associated a- and (3- subunit heterodimers. There are 16 different a- subunits and 8 different [3- subunits which associate in various combinations 10 to form more than 20 different heterodimers (Hynes 1992). Integrins generally contain large extracellular domains formed by a (~ 1000 residues) and 0 (~ 750 residues) subunits and relatively short cytoplasmic domains with the exception of 04 integrin, which has a cytoplasmic domain with >1000 amino acid residues (Hogervorst et al., 1990; Suzuki and Naitoh, 1990; Tamura et al, 1990). While the extracellular domains interact with ECM proteins, the cytoplasmic domains interact with cytoskeletal elements. Thus, integrins have been established as cell surface receptors which can mediate the bidirectional transfer of information. In "outside-in" signaling, integrin interacts with the cytoskeleton and signaling molecules to generate cytoskeletal rearrangements as well as secondary signals. In "inside-out" signaling, changes in the cytoskeleton influence the integrin-ligand binding affinity and thereby regulate integrin function (Crowe et al., 1994; Shimizu et al., 1990). In the mammary gland, many of the influences of B M on epithelial morphogensis and differentiation can be recapitulated by laminin or can be blocked by anti-laminin antibodies (Roskelley et al., 1994). The cell surface integrin receptors that interact with laminin include a l , a2, a3, a6, 01 and 04 integrins (Koukouhs et al., 1991; Pignatelli et al., 1991; Howlett et al., 1995). While all of these subunits are present in the mammary epithelial cells, cc6 only interacts with 04, and blocking antibodies against a604 prevents 0-casein induction (Muschler et al., 1999) and morphogenesis (Weaver et al., 1997). 1.6. The cytoskeleton a. Microfilaments (actin) In electron micrographs actin filaments have a diameter of ~ 8 nm. They consist of a tight helix of uniformly oriented globular actin molecules. The filaments are polarized 11 with a slow growing minus end and a fast growing plus end. The dynamics of actin assembly/disassembly plays a major role in determining cellular functions such as change of cell shape and cell migration. The most studied actin containing structures include focal adhesions, stress fibers, lamellipodia and filopodia (Small et al., 1999). Lamellipodia consists of a thin wave-like extensions of cells with a planar meshwork. of actin filaments whose fast growing ends are oriented outwards. Filopodia, consist of long extensions of cell membrane containing actin filaments. Focal adhesions, stress fibers and peripheral actin bundles are located at the edge of the cells or at the edge of non-motile lamellipodia. These latter structures require substrate attachment for their formation. Formation of these actin-based structures is also regulated by Rho family of G proteins (Hall et al, 1998). Stress fiber and focal contact formation is induced by Rho, while lamellipodia and filapodia formation are induced by Rac and Cdc42, respectively (Nobes and Hall, 1995; Hall, 1998). Rho, Rac and Cdc42 have also been implicated in the formation of interactions with ECM, but the extent of these interactions are still unclear (Nobes at el., 1995; Clark et al., 1998). However, in my work I was interested in actin associated junctions. In particular, I was interested in E-cadherin-mediated adherens junctions and occludin-mediated tight junctions. Pharmacological agents that stabilize or destabilize actin filaments provide important tools to study the actin dynamics in cells. Cyochalasin D (CD) is a fungal metabolite that binds to the plus end of actin filaments and prevents actin polymerization. Thus, CD reduces the actin filament mass and also kinetically stabilizes actin dynamics at the plus end (Cooper, 1987). 12 b. Intermediate filaments (IF) In contrast to microfilaments and microtubules, whose components are highly conserved throughout evolution and are very similar within cells of a given species, intermediate filaments are more diverse in number, sequence and abundance (Fuchs and Weber, 1994). Despite their diversity, members of the IF super family share a common structure. Two parallel a-helical chains intertwine to form a dimer, and four dimers associate in an anti-parallel fashion to form protofibrils. Three to four protofibrils intertwine to produce a intermediate filament with a diameter of 10 nm. IFs can be classified into four types and epithelial cells generally express type I and type II keratin filaments at 1:1 ratio. Type III IFs are composed of vimentin and are mainly observed in mesenchymal cells. Type IV IFs are composed of neurofilaments and are expressed in neurons. The IFs and associated proteins form filamentous networks, that are thought to provide integrity and strength throughout the tissue. IFs have been shown to reversibly link to the plasmamembrane and to other cytoskeletal components to maintain cell shape and withstand mechanical stress (Houseweart and Cleveland, 1998) as well as to position the nucleus within the cell. Plectin is an IF-associated protein which, along with desmoplakins, links the IF network to desmosomes and thereby distributes the mechanical stress to the whole tissue. Similarly, IF linkage to hemidesmosomes facilitates cell attachment to the basement membrane and it initiates cell polarity. Identification of the neurotoxin acrylamide (Ac) as a pharmacological agent that can reversibly collapse (Eckert, 1985) the intermediate filament network in epithelial cells has provided us with a tool for studying IF function. 13 c. Microtubules Microtubules are formed by the polymerization of afi-tubulin dimers into a small polymer 'nucleus', followed by elongation of the polymer at each end by reversible and noncovalent addition of subunits. The plus end shows more dynamic instability than the minus end of the filament. Net growth at the plus end leads to elongation at the plus end and shortening at the minus end. Microtubules appear as long, fairly straight, unbranched filaments about 25nm in diameter (Alberts et al, 1994). They are present in virtually all cell types with the exception of red blood cells. Microtublues play major roles in cell division, organelle transport (Allan et al., 1991), and secretory vesicle and granule transport (Kreis et al., 1989; Tooze and Burke, 1987). The effects of anti-microtubule drugs have provided tools for examining the roles of microtubules in cellular functions. Treatment of mammary epithelial cells with colchicine leads to inhibition of protein secretion in vivo (Patton, 1974) and in vitro (Knudson et al., 1978). Nocodazole (ND), another pharmacological microthbule disrupting agent, also displays similar effects on the lactating mammary epithelium. ND disrupts the microtubule network completely by depolymerizing the filaments in a reversible manner and releasing tubulin into the soluble form (Rennison et al., 1992). The three major cytoskeletal components can interact with each other through common linker proteins. For example, Rho family G proteins that regulate actin structures also bind directly to the microtubules (Best et al., 1996) and to the kinesin-associated protein kinectin (Hotta et al., 1996) giving an indication of actin-microtubule interaction. The tumor supressor protein adenomatous polyposis (APC) is also known to, 14 bind to microtubules and link to the actin cytoskeleton via a-catenin (Munemitsu et al., 1994). Keratin filaments interact with both microtubules and actin filaments to form cross-bridges via the linker protein plectin (Houseweart and Cleveland, 1998) while BPAG1 links keratin to actin (Yang et al, 1996). Interactions between each of the cytoskeletal components are important in maintaining the balance of forces in a cell. Recent studies have revealed that the balance of mechanical tension generated through molecular interaction within the cytoskeleton, as well as cell-cell and cell-ECM interaction, provides information for cell stability and structure. Thus, cytoskeletal interactions play a major role in maintaining cell polarity and tissue shape. 1.7. Cell-Cell and Cell-ECM Junctions Epithelia are made up of contiguous cellular sheets which act as barriers between distinct biological compartments. A major function of any given epithelium is to selectively regulate the passage of ions, fluids and organic molecules between these compartments in a polarized fashion. The basal epithelial cell surface is attached to a specialized extracellular matrix, the basal lamina, which abuts upon the connective tissue compartment (Simons and Fuller, 1985). The apical surface is often free, and in the case of glandular epithelia, faces a central lumen. The lateral surfaces of the epithelial cells interact with each other via specialized intracellular junctions. This junctional complex consists of desmosomes, adherens junctions and tight junctions (Fig 1). Usually the most basally located junction is the desmosome, which looks like a "spot weld" in electron micrographs. This disk shaped spot is a protein plaque to which intermediate filaments, usually cytokeratin are attached on the cytoplasmic side. At the surface of one cell, this 15 disk structure is matched with an identical structure of the adjacent cell membrane. Apical to the desmosome is the adherens junction which encircles the cell and provides adhesion of one cell to its neighbors via cell surface cadherins (Rajasekaran et al., 1996). The tight junction (TJ), also known as zonulae occludentes, is the most apical of all the junctions. Zonula refers to the junction forming a band that completely encircles the cell. Occludentes refers to membrane fusion closing off the intracellular space (Gumbiner, 1987). In addition to cell-cell junctions, epithelial cells also form cell-ECM junctions, one of which is the hemidesmosome (Dowling et al., 1996). a. Tight Junctions (TJ). A major breakthrough in the molecular characterization of the TJ was the identification of the TJ associated protein ZO-1 (Stevenson et al., 1986). This was achieved by generating specific mAbs TJ-enriched mouse liver fractions followed by the identification of specific antigens to which these mAbs bind. In 1991, a 220 kDa human protein was characterized and identified as the mouse homologue of ZO-1 (Itoh et al., 1993). Biochemical analysis revealed that ZO-1 is a cytoplasmic peripheral membrane phosphoprotein. The actin cytoskeleton links to the tight junctions by directly interacting with ZO-1 (Itoh et al., 1997).The amount of ZO-1 is not directly related to the junctional permeability but rather it is directly related to the number of fibrils present in the TJ network (Stevenson et al, 1988). A few years after the identification of ZO-1, cingulin, another TJ associated protein, was identified and characterized. Cingulin is also a cytoplasmic protein and immunolocalization studies revealed that it is about three times farther from the TJ membrane than ZO-1. ZO-2 and 7H6 are two other proteins that have 16 been identified and characterized (Citi, 1993). ZO-2 was first identified as a 160 kDa cytoplasmic peripheral membrane protein that co-immunoprecipitates with, and therefore physically associates with, ZO-1. Occludin is a transmembrane protein that is exclusively localized to the TJ at immunofluoresence and immunoelectron microscopic levels and it is the first transmembrane tight junction protein to be identified (Fig. 1; Furuse et al., 1993). cDNA sequence analysis suggests that occludin is composed of four transmembrane domains, a long carboxyl-terminal, a short amino-terminal domain and two extracellular loops (Furuse et al., 1996). Double immunofluorescence staining of cultured human intestinal epithelial cells with mAbs against ZO-1 and occludin show that they are precisely co-localized (Ando-Akatsuka et al., 1996). Immunofluorescence and laser scan microscopy reveals that when full length occludin is introduced into epithelial cells, it is correctly delivered to the membrane and is incorporated into existing TJ. Furthermore, deletion mutation and fusion protein analysis reveals that binding of ZO-1 directly to the cytoplasmic domain of occludin appears to be required for the localization of occludin to the TJ (Furuse et al., 1994). ZO-3 is a more recently identified cytoplasmic protein that interacts with ZO-1 and occludin in the formation of TJ (Haskins et al., 1998). Claudin-1 and claudin-2 are two 23 kDa proteins identified as second and third transmembrane components of TJ (Furuse et al., 1998). When these proteins are introduced into cultured epithelia cells (MDCK), they are properly incorporated into the pre-existing tight junctions (Morita et al, 1999). Interestingly, these two proteins have no sequence similarity to occludin. 17 b. Adherens Junctions E-cadherin is a member of the cadherin family that mediates homophilic calcium-dependent cell-cell interactions to maintain the normal epithelial phenotype by initiating the formation of adherens junctions (AJ; Nelson et al, 1990). The cytoplasmic domain of E-cadherin interacts with catenin family molecules which are comprised of three members; a-, fi- and y-catenin. Adherens junctions also interact with the actin cytoskeleton via a-catenins (Fig. 5; Knudsen et al., 1995). The fully functional adhesive properties of E-cadherin depend on the integrity of these cadherin-catenin-actin complexes, and loss of this integrity can lead to loss of differentiation and gain of increased motility and invasive properties. Several studies have demonstrated a strong correlation between loss of E-cadherin-catenin expression and both loss of epithelial phenotype and increased invasive phenotype (Rasbridge et al, 1993; Shiozaki et al., 1991). More interestingly, the forced expression of E-cadherin in several tumor cell lines and epithelial cells having fibroblastic-like morphology re-establish the epithelial phenotype and decrease the invasive phenotype (Frixen et al., 1991; Chen et al., 1997). c. Desmosomes The desmosome is a cell-cell junction that is localized basal to adherens junctions, and it is composed of a symmetrical membranous plaque. Each half of the desmosome is derived from an adjacent cell and contains a membranous protein plaque to where IF are attached (Kelly, 1966). The adhesion molecules at the junction belongs to the transmembrane glycoprotein, cadherin superfamily and they are called desmogleins and desmocolins (Fig. 1; Koch et al, 1990; Buxton et al., 1993). Like cadherin-mediated 18 adherens junctions, desmosomes also contain a member of the catenin family, plakoglobin (Gumbiner and McCrea, 1993). Desmoplakin I and II are two other cytoplasmic (Green et al, 1988) proteins associated with the desmosomal protein plaque. The carboxy terminals of desmoplakin proteins interact with amino terminal head groups of keratin filaments to link the desmosomes to the intermediate filament cytoskeleton (Kouklis et al., 1994), and this interaction is essential for the assembly of desmosomes (Galhcano et al., 1998). d. Hemidesmosome Hemidesmosomes are punctate junctions connecting the basal surface of epithelial cells to the basement membrane. At the ultrastructural level they look like tripartite structures with an electron dense juxtra-membranous plaque and an innermost plaque which links to the keratin cytoskeleton (Shienvold and Kelly, 1976). At the molecular level, this junction consists of a link between the cell surface integrin a6pM and the basement membrane protein laminin. In addition to a6(34 integrin, five other hemidesmosomal plaque molecules that have been identified. These molecules are plectin (Hieda et al., 1992), bullous pemphigoid antigen 1 (BPAG 1; Sawamura et al., 1991), BPAG 2 (McGrath et al, 1995), 6A5 antigen (Kurpakus et al., 1991) and HD-1 protein (Hieda et al., 1992). BPAG 2 is known to interact with the cytoplasmic domain of the oc6 integrin subunit and it may be involved in cell adhesion to the basement membrane. The rest of the hemidesmosome-associated proteins are localized to the inner plaque. BPAG 1 contains an actin-binding domain in its N H 2 terminus and a keratin filament binding domain at COOH terminus (Leung et al, 1999). Thus, BPAG 1 is involved in linking the 19 innermost plaque to the cytoskeleton (Fig. 2). BPAG 1 knock-out mice lack the inner plaque and are unable to form connections with the keratin cytoskeleton (Guo et al, 1995). The large cytoplasmic domain of [34 integrin contains two pairs of type III fibronectin (Fn) like modules linked by a 142 amino acid linker region. The first Fn-like domain and the connecting region is known to be essential for the assembly of hemidesmosomes (Spinardi et al, 1993). [34 integrin null mice demonstrated extensive detachment of the epidermis and other squamous epithelia along with neonatal death (van der Neut et al., 1996). This dramatically reduced adhesion is accompanied by the absence of hemidesmosomes. Thus, a6pM-larninin interactions provide an essential link between epithelial cells and the basement membrane and the also initiate apical/basal cell polarity due to their linkage to the actin and intermediate filament cytoskeleton. 20 Fig 1: Junct ional complexes in epithelial cel ls . In the epithelial cells there are three major cell-cell junctional complexes. The most apically located tight junctions consist of the transmembrane protein occludin and a group of cytoplasmic proteins. These cytoplasmic proteins include ZO-1, which directly interacts with occludin, ZO-2 and the actin filaments. 7H6 and cingulin are two other proteins that are associated with the cytoplasmic protein plaque. Basal to the tight junction is the adherens junction. The cytoplasmic domain of the adherens junction protein E-cadherin interacts with P-catenin. P-Catenin also interacts with oc-catenin. oc-catenin interacts with a-actinin which links filamentous actin to the adherens junction. The most basalfy located junction complex is the desmosome. Desmoglein (DL) and desmocolin (DC), two transmembrane proteins in the desmosome belong to the cadherin super-family of glycoproteins. The cytoplasmic domains on the transmembrane proteins are associated with plakoglobin (P), which is a catenin family protein. Desmoplakin I (DP I) and DP II are two cytoplasmic proteins that interact with plakoglobin and with keratin filaments, thus linking the intermediate filament cytoskeleton to desmosomes. 21 12 Fig. 2: Hemidesmosomes Upon oc6f34 integrin binding to laminin, hemidesmosomes are assembled. BPAG 2 interacts with the cytoplasmic domain of oc6. BPAG 1 interacts with HD-1, 6A5, and plectin to form a protein complex that interacts with the (34 integrin subunit as well with as the keratin intermediate filament network. 23 1.8. Hypothesis A flexible extracellular matrix allows mammary epithelial cells to undergo morphogenisis and express tissue specific milk protein genes. As discussed previously, these differentiating cells require changes in cell organization, polarity, and cell-cell and cell-matrix interactions. It is also known that the actin cytoskeleton is associated with cell-rounding, cell-cell adherens junctions and tight junctions. Where as the keratin intermediate filaments are associated with the initiation of cell polarity by forming hemidesmosomes. Hemidesmosome-associated a6(34 integrin is important in integrin-dependent (3-casein gene induction (Fig 3). Thus, I hypothesize that the effects of basement membrane mediated mammary epithelial morphogenisis and differentiation should require a functional cytoskeleton. If my hypothesis is true, disruption of the cytoskeleton should prevent morphogenesis and differentiation as indicated by the inhibition of differentiation specific milk protein gene expression. To test this hypothesis, I will disrupt microtubule, keratin and actin cytoskeletal components with pharmacological agents ND, Ac and CD and will assay for the ability for the mammary epithelial cells to form three-dimensional alveolar-like structures and induce milk protein genes, lactoferrin and (3-casein. 25 Fig 3: Mouse mammary epithelial cell in culture i) When mouse epithelial cells are placed on tissue culture plastic dishes, cells proliferate and grow into an unpolarized flat monolayer, covering the complete surface area of the dish. ii) . When mouse mammary epithelial cells are placed onto tissue culture plastic dishes coated with a reconstituted basement membrane gel (Matrigel), cells start to aggregate within hours after plating. These cells pull the matrix around them and undergo remodeling of the aggregates. This reorganization results in structures similar to alveoli in the mammary gland. These structures secrete milk proteins into the central lumen. 26 II Tissue. C^l+vwe PlOusKc-Cell rounding -actin Lactoferrin HaVx^ei -cell rounding Adherens junctions -actin Tight junctions -actin Hemidesmosomes -keratin Cell polarity -keratin-a6P4 integrin P-Casein induction -a604 integrin 2 7 II. MATERIALS AND METHODS 1. Cell line The functional mammary epithelial cell strain COMMA-ID is a cell strain that was initially derived from the mammary-alveolar epithelium of a mid pregnant mouse (Danielson et al., 1984). In the presence of exogenously added extracellular matrix some of the COMMA-ID heterogeneous cells respond to lactogenic hormones and are induced to express the milk protein P-casein. CID-1 is a population of cells that originated from P-casein expressing cells of COMMA-ID strain (Schmidhauser et al., 1990). Desprez et al used limited dilution techniques to isolate the scp2 cell clone from the CID-1 cell population. Scp2 cells are composed of small cuboidal epithelial cells that grow as flat monolayers on tissue culture plastic and absolutely require exogenously-added ECM to undergo morphogenesis and induce P-casein expression (Desprez et al, 1993). 2. Cell culture Scp2 cells were routinely cultured in Dulbecco's modified Eagle's medium (DMEM)/F12 medium (1:1, Sigma, St. Louis, MO) supplemented with 5% fetal bovine serum (FBS, Hyclone, Logan, UT) and insulin (5|ig/ml, Sigma), at 37°C in a humidified atmosphere of 5% CO2 in air. The antibiotic gentamycin was added at 50|!g/ml final concentration in the media to prevent bacterial contamination. Cells were seeded at approximately 1.5 X 106 cells / 100 mm tissue culture plastic dishes and the culture medium was changed every 48 hrs. When the cells were approximately 95% confluent, cells were trypsinized (.5% 28 trypsin (StemCell)/ .06% EDTA in Ca z + /Mg z + free hanks balanced salt solution (HBSS, Sigma)) and removed from the tissue culture dish. Cell stocks were maintained in DMEM/F12 media containing 20% dimethyl sulfoxide (DMSO) and 25% FBS, and were stored in liquid N 2 (-196°C). 2.1. Maintenance of a pure mammary epithelial cell population During cell culture, epithelial cells can transform into cells with fibroblastic characteristics, and these cells have to be removed in order to maintain a pure epithelial population. Throughout these experiments, a differential plating technique was used to maintain a pure mammary epithelial cell culture (Somasiri and Roskelley, 1999). Cells with fibroblastic characteristics adhere and spread on tissue culture plastic faster than epithelial cells. Therefore, when cells were trypsinized, plated on tissue culture dishes and placed in the incubator for 20 min, all the unattached cells were transferred to new tissue culture dishes. In this process fibroblastic cells remain in the original dish, and the epithelial cells were transferred to the new dish. In a pure epithelial culture, 99% of the cells maintain cobblestone epithelial morphology. 2.2. MATRIGEL® Basement Membrane Matrix (EHS) MATRIGEL (Collaborative Res., Bedford MA) Basement Membrane is a solubilized basement membrane extracted from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma, a tumor rich in extracellular matrix proteins. The major component is laminin, followed by collagen IV, heparan sulfate proteoglycans and entactin. It also contains TGF-p\ fibroblast growth factor, tissue plasminogen activator and other naturally occurring growth factors of EHS tumor. 29 2.3. Preparation of MATRIGEL coated dishes Matrigel stocks were thawed overnight on ice, in a 4°C fridge and diluted 1:1 with DMEM/F12 base media. Matrigel solutions were always kept on ice to prevent premature gelling. 35 mm tissue culture dishes were chilled on ice and 125)0.1 of diluted matrigel solution was added to the center of each dish. Matrigel was spread evenly on the entire surface of the dish and incubated at 37°C for at least an hour. For immunofluorescent staining, cells had to be cultured on glass cover-slips. Thus, cover-slips were sterilized by immersing in 70% ethanol, and then placed vertically in twelve well tissue culture plates that were placed in a tissue culture hood. When the EtOH had completely evaporated and cover glasses were dry, the plates were shaken to drop each cover glass to the bottom of each well. The twelve well plates were chilled on ice and 75(il of diluted Matrigel solution was placed on the cover-slip and spread evenly. The plates were incubated at 37°C for at least an hour for gelling. 2.4. Differentiation assay Scp2 monolayers were trypsinized and cells were re-suspended in DMEM/F12 media containing 1% FBS, insulin and gentamysin. Cells were seeded at 5 X 10 6 / 35 mm onto the matrigel gels in 35 mm tissue culture dishes. The cultures were incubated at 37°C, 5% CO2 and 95% air and observed daily. The medium was changed to the differentiation media 18 hours after plating. This medium was composed of serum free DMEM/F12 media supplemented with insulin (5(ig/ml), prolactin (3ug/ml) and hydrocortisone (l|ig/ml). The medium was changed every 48 hours and cells were collected for protein analysis or for immunostaining at 24, 48 and 72 hour time points. 30 3. Cytoskeletal disruption assays 3.1. Monolayer cultures a. Microtubule disruption assay Cells were cultured on glass cover-slips to -95% confluency and the media was changed to growth media containing ND (0-100|ig/ml). ND stock solution was made at lp:g/|il in DMSO and was stored at -20°C. Just prior to adding to the cells, stock solution was diluted in growth media. Cells were fixed every two hours for 24 hours and stained for tubulin. b. Keratin disruption assay Cells were cultured on glass cover-slips to -95% confluency and the media was changed to growth media containing Ac (0-80mM). Ac stock solution was made at 400mM in DMEM/F12 and stored at -20°C. Just prior to adding to the cells, stock solution was diluted in growth media. Cells were fixed every two hours for 24 hours and stained for cytokeratin. c. Actin disruption assay Cells were cultured on glass cover-slips to -95% confluency and the media was changed to growth media containing CD (1-lOu.g/ml). CD (Sigma) was dissolved in DMSO to make a stock solution of lOug/pJ and stored at -20°C. Just prior to adding to 31 the cells, stock solution was diluted in growth media. Cells were fixed every two hours for 24 hours and stained for actin. 3.2. Cells cultured on Matrigel a. Microtubule disruption assay Cells were placed on Matrigel-coated 35mm tissue culture dishes and glass coverslips as described above. 18 hours after plating, the media was changed to freshly made differentiation medium that contained the microtubule disrupter ND. Just prior to adding to the cells, stock solution was diluted in DMEM/F12, supplemented with hydrocortisone, prolactin and insulin, to achieve a final ND concentration series of 0 p:g/ml, 2 pig/ml, 9 uvg/ml, 30 ug/ml and 100 u\g/ml. After 4hr of incubation at 37°C, cells were rinsed three times with DMEM/F12 base media to completely remove any remaining ND. Fresh differentiation media was added and media was changed every 48 hours. The cells on 35 mm dishes were collected after 72 hours of recovery and assayed for milk proteins by western blot analysis. The cells on the cover-slips were fixed and assayed for microtubule cytoskeleton by immunostaining. b. Keratin disruption assay Cells were placed on Matrigel-coated 35mm tissue culture dishes and glass coverslips as described above. 18 hours after plating, the media was changed to freshly made differentiation medium that contained the keratin disrupter Ac. Just prior to adding to the cells, stock solution was diluted in DMEM/F12, supplemented with hydrocortisone, prolactin and insulin, to achieve a final Ac concentration series of OmM, 32 lOmM, 20mM, 40mM and 80mM. After 4hour of incubation at 37 C, cells were rinsed three times with DMEM/F12 base media to completely remove any remaining Ac. Fresh differentiation media was added and media was changed every 48 hours. The cells on 35 mm dishes were collected after 72 hours of recovery and assayed for milk proteins by western blot analysis. The cells on the cover-slips were fixed and assayed for keratin cytoskeleton by immunostaining. c. Actin disruption assay Cells were placed on Matrigel-coated 35mm tissue culture dishes and glass coverslips as described above. 18 hours after plating, the media was changed to freshly made differentiation media that contained the actin disrupter CD. Just prior to adding to the cells, stock solution was diluted in DMEM/F12, supplemented with hydrocortisone, prolactin and insulin, to achieve a CD final concentration series of 0 u.g/ml, .2 [ig/ml, .9 fig/ml, 3 (ig/ml and 10 |ig/ml. After 4hr of incubation at 37°C, cells were rinsed three times with DMEM/F12 base media to completely remove any remaining cytochalasis. Fresh differentiation media was added and media was changed every 48 hours. The cells on 35 mm dishes were collected alter 72 hours of recovery and assayed for milk proteins by western blot analysis. The cells on the cover-slips were fixed and assayed for actin cytoskeleton by immunostaining. To assay for the ability of the cells to completely recover the actin cytoskeleton and undergo morphogenesis, cells were treated with 10p:g/ml CD for 4 hours and were recovered in differentiation media for 6 days. Samples of cells were collected every 24 hours and were assayed for P-casein and lactoferrin by Western blotting. Actin and 33 occludin localization were assayed by staining with rhodamine-labeled phallotoxin or with indirect immunostaing using anti-occludin antibody. 4. Western blot analysis 4.1. For P-casein and lactoferrin Cells that were cultured on Matrigel were rinsed once with DMEM/F12 base media. 1 ml of Dispase (Collaborative Research) was added to each 35 mm dish and incubated at 37°C for 45 min. Dispase digests the ECM proteins and releases the cell clusters. Cells were collected to 1.5ml centrifuge tubes and washed three times with DMEM/F12 (cells were centrifuged at 3000g for 4 min and the cell pellet was completely resuspended in each cycle). To the cell pellet, 75pl of RIPA buffer (150 mM NaCl, 50 mM Tris pH 7.4, 5 mM EDTA, 5% NP-40, 1% sodium deoxycholate (DOC), and 0.1% sodium dodecyl sulphate (SDS), aprotinin, leupeptin, phenylmethyl-sulfonyl fluoride (PMSF)) was added and incubated on ice for 10 min. Tubes were vortexed for 10 sec to assist lysis before centrifuging at 13,000g in a microcentrifuge placed at 4 °C. Supernatants were collected and stored at -70°C. Protein concentrations were measured using a BioRad protein assay kit. Lysates were normalyzed for equal amount of protein by adding RIPA buffer and sample buffer (20% glycerol, . IM Tris pH 6.8, 4% SDS, .004% bromo phenol blue, 10% 2-mercaptoethanol). 20p:g of was protein separated on 1.5 mm thick 13% SDS-poly acrylamide gel electrophoresis (PAGE) gels and transferred on to PVDF (BioRad) membranes. Membranes were incubated for 12 hr with 4% bovine serum albumin (BSA) and 5% FBS in Tris buffered saline (TBS)-Tween-20 to block the nonspecific sites. Milk proteins were detected by probing with mouse monoclonal anti-p-34 casein antibody (from Dr. C. Kaetzel, Institute of Pathology, Case Western Reserve University, Cleveland, OH) or with affinity-purified polyclonal rabbit anti-lactoferrin antibody (Teng et al., 1989) overnight at 4°C. Unbound primary antibodies were removed by washing 3 X 5 min each and 3 X 10 min each with TBS-T, followed by incubation with HRP-conjugated anti-mouse IgG for 1 hr. Bands were visualized with enhanced chemiluminescent (ECL, Amersham, Arlington Heights, IL) reagents followed by autoradiography. 4.2. For tight junction protein occludin Since, occludin phosphorylation is reported with tight junction formation, I wanted to assay for possible phosphorylation events and occludin localization in the mammary epithelial cells. I examined Scp2 monolayers, cells that were rounded in the absence of Matrigel for occludin expression. For Matrigel independent cell rounding, scp2 cells were plated on polyHema coated 100mm tissue culture dishes at 20X106 cells per dish in DMEM/F12 media supplemented with 1% FBS, insulin and hydrocortisone. PolyHEMA is a nonadhesive substratum that removes the negative charge from tissue culture dishes to prevent cell attachment to the dish. As a result of inability to adhere cells remained rounded and formed cell aggregates. The cells were allowed to round up and aggregate for 3 days and were then collected for protein analysis. The MDCK cell line was used as a positive control. Collected cell pellets were washed 3 times with ice-cold PBS, and the 300(il of ice-cold NP-40 lysis buffer (25 mM Hepes/NaOH pH 7.4, 150 mM Nacl, 4mM EDTA, 25 mM NaF, 1% NP-40, 1 mM Na 3 V0 4 , 1 mM PMSF, 10|ig/ml leupeptin, 10 |!g/ml aprotinin) was added to each. NP-40 lysis buffer was added 35 on to MDCK and Scp2 cell monolayers that were rinsed with cold phosphate buffered saline (PBS) and were then scraped and collected into 1.5 ml microcentrifuge tubes. All the tubes were gently agitated on a shaker for 30min at 4°C. After centrifugation at 10,000g for 30 min, supernatants were collected as NP-40 soluble fractions. 75 (il of SDS lysis buffer (25 mM Hepes/NaOH pH 7.5, 4 mM EDTA, 25mM NaF, 1% SDS and 1 mM Na3V04) was added to each pellet and hormogenized using a .5 ml microcentrifuge tube attached to a glass rod. Walls of each .5 ml microcentrifuge tube were washed with 75 p:l of NP-40 lysis buffer into the hormogente. The lysates were subjected to four rounds of sonications (10 sec sonication, 5 sec vortex and 10 sec on ice) and then centrifuged at 10,000g for 30 min and the supernatants were collected as cytoskeletal fractions. Protein concentrations of each sample were determined by BioRad protein assay and 20 p:g of protein from each sample was resolved by one-dimensional SDS-PAGE electrophoresis. Proteins were transferred on to a PVDF membrane, blocked overnight at 4°C in the blocking solution (5% FBS, 4% BSA, 2.5% normal goat serum in TBS-T), and incubated with monoclonal mouse anti-occludin (clone OC-3F10; 1 p:g/ml in 1% BSA in TBS-T). The primary antibody was detected with HRP conjugated rabbit IgG (1: 5000 in TBS-T), followed by ECL reaction and autoradiography. 5. Immunofluorescence Microscopy It is not possible to use conventional light microscopy techniques to examine individual cytoskeletal components. Therefore, indirect immunofluorescence staining and confocal microscopy techniques were used to identify actin, keratin and microtubule cytoskeketal components as well as the cell-cell junction proteins. Cells were maintained 36 as monolayers or in matrigel on glass cover-slips before treatment with each disrupting agent. After 4hr treatment or 72hr recovery, cells were fixed and stained for respective cytoskeletal component. 5.1. Tubulin staining Cells were rinsed once with PBS and then fixed with 3.7% paraformaldehyde in PBS at pH 7.4 for 15 min at room temperature. After fixing, cells were kept in 1% paraformaldehyde at 4°C for up to 2 weeks before staining. Fixed cells were rinsed in PBS and blocked with 10% normal goat serum (NGS), 1% BSA in PBS for 20 min. Cells were rinsed 3 times 5 min each and incubated with mouse anti a-tubulin antibody (1/20; CalBiochem, CA.) in 1% BSA for lhr at room temperature. For the controls, mouse IgG was used in instead of the primary antibody. Cells were rinsed 3 X 5min each and incubated with the secondary antibody, FITC-conjugated anti mouse IgG. Then rinsed 3 times 5min each with PBS and cover-slips were mounted on to glass slides with 90% glycerol and sealed with nail polish. 5.2. Keratin staining Cells were rinsed with PBS and fixed with methanol/acetone (1:1) for 5 min at -20°C. Fixed cells were stored in PBS at 4°C for up to two weeks. Cells were permeabilized with 1% Tryton-X 100 in PBS for 5 min and blocked with 10% NGS, 1% BSA in PBS for 20 min, and then incubated with Rabbit polyclonal anti-keratin antibody (1/400; DAKO, Carpintera, CA) in 1% BSA in PBS for 1 hr, rinsed 3 times 5min each with PBS and 37 incubated with FITC conjugated rabbit IgG for 1 hr. Cover-slips were rinsed and mounted on glass slides with 90% glycerol. 5.3. Actin staining Cells were rinsed with PBS, fixed with 3.7% paraformaldehyde in PBS for 15 min at room temperature, extracted with acetone at -20°C for 5min, and then air-dried. Cells were incubated with rhodamine-labelled phallotoxin (Molecular Probes, Eugene, OR) in PBS for 20 min, rinsed 2 X 5 min each, and then the coverslips were mounted on glass slides with 90% glycerol. 5.4. E-cadherin staining Cells were fixed with -20° C methanol for 15 min, rinsed with PBS, incubated with the blocking solution (10% NGS / 1% BSA in PBS) for 20 min at room temperature, and incubated with antibodies against mouse anti-E-cadherin (Transduction laboratories). Primary antibody binding was detected by incubating for 1 hr at room temperature with FITC-conjugated AffiniPure goat anti mouse IgG (Jackson ImmunoResearch 1:100). The Cover-slips were rinsed 3 X 5 min each with PBS and mounted on glass slides with 90% glycerol. 5.5. Occludin staining Cells were fixed with 1% paraformaldehyde in PBS for 20 min at room temperature, rinsed with PBS and permeabilized with 0.2% Triton X-100 in PBS for 20 min. After rinsing the cells in PBS 3 X 5 min each, they were blocked with 10% NGS, 1% BSA in 38 PBS for 20 min and incubated with mouse anti-occludin antibody (1.5|ig/ml) for 1 hr at room temperature. Cells were rinsed and incubated with FITC-conjugated anti-mouse IgG for 1 hr before mounting onto glass slides with 90% glycerol. 5.6. Image collection Phase pictures of the cells were collected using a Nikon F camera connected to a Nikon TMS microscope unit. Immunofluoresence images were collected using a BioRad MRC 600 confocal unit connected to a Nikon optiphot-2 microscope or directly photographed using a Zeiss axiophot microscope onto T-Max 400 film. The negatives were then digitized by scaning on a poloroid sprint scan plus film and slide scanner. 39 III. RESULTS 1. Optimal conditions to disrupt cytoskeletal components: microtubules, keratin and actin The experiments in this thesis were dependent on my ability to disrupt the three major cytoskeletal elements in mammary epithelial cells. Therefore, my initial goal was to optimize the conditions for complete disruption of major cytoskeletal elements. I used the pharmacological agents ND, Ac and CD to disrupt microtubules, keratin filaments and actin filaments respectively. In an attempt to minimize the nonspecific effects on the cells, initial studies were designed to identify both the lowest concentrations of each disrupting agent and the shortest time intervals required to completely disrupt each cytoskeletal component. When plated on tissue culture plastic, scp2 cells grow as epithelial monolayers with the classical 'cobblestone' epithelial morphology. Under these conditions the cells form close contacts with their neighbors which are mediated by adherens junctions (Somasiri et al., 1999; Lochter et al, 1997). Thus, treatment of these monolayers with disrupting agents provided a good assay to initially examine the effects of each agent on the cytoskeleton and on the cell morphology (data summarized in Table 1). 1.1. Microtubule disruption In the cell monolayers treated with vehicle alone (control), the microtubules formed an unorganized filamentous network that radiated throughout the cell cytoplasm (Fig. 4i). Monolayers were treated with the microtubule disrupting agent ND at concentrations 40 ranging from 0 to 10|ig/ml for times ranging from 1 hour to 24 hours. At lower concentrations (0 to 8|ig/ml) and shorter times (0 to 2 hours) the disruption was incomplete and some remaining filaments were seen. Treatment with higher concentrations (> 10|ig/ml) and longer times (> 4 hours) led to the complete loss of microtubule filament network and staining was diffused throughout the cytoplasm of each cell. Treatment with >10u\g/ml ND, had similar effects as treatment with 10p.g/ml ND. Immunofluoresence staining of tubulin revealed that the microtubules could be most effectively and optimally disrupted with a treatment of lOug/ml ND for 4 hours. With this treatment regime tubulin was no longer found in filaments. Instead, it was uniformly distributed throughout the cytoplasm of each cell (Fig 4ii). The unstained clear areas represented cell nuclei (Fig 4ii). This was confirmed using DNA binding dye 4'-6-diamidine-2-phenyl indole (DAPI). Phase microscopy of live cells indicated that there was no difference in morphology between monolayers treated with optimal conditions of ND and untreated cell monolayers (Fig. 5). Thus, cells maintained both a flattened cell shape and cell-cell interactions in both conditions. 1.2. Intermediate filament disruption In the cell monolayers treated with vehicle alone (control), the keratin intermediate filaments formed networks that caged nuclei of the cells and extended to the cell periphery (Fig. 4iii). Monolayers were treated with the cytokeratin disrupting agent Ac at concentrations ranging from 0 to 40 mM for times ranging from 1 hour to 24 hours. At lower concentrations (0 to 5 mM) and shorter times (0 to 2 hours) the majority of the 41 keratin filaments were unaffected. Treatment with higher doses (> 2()mM) and longer times (> 4 hours) led to disruption of the keratin filaments. They were partially collapsed, shortened and the staining intensity was increased. Immunofluoresence staining revealed that the keratin cytoskeleton could be effectively and optimally disrupted by a treatment of 20mM Ac for 4 hrs (Fig 4iv). Phase microscopy of live cells indicated that the optimal treatment with Ac caused the cells to retract slightly such that small spaces formed between the cells (Fig 5). Thus, the integrity of the monolayer was disrupted which suggests that cell-cell junctions were affected. 1.3. Actin filament disruption In the cell monolayers treated with vehicle alone (control), the actin filaments were localized as sub-plasmalemmal cortical rings (Fig 4v). In monolayers, this actin co-localizes with the adherens junction proteins E-cadherin and P-catenin (Somasiri et al., 1999). There were almost no visible actin stress fibers radiating across the cytoplasm of these epithelial cells. Monolayers were treated with the f-actin disrupting agent CD at concentrations ranging from 0 to 10p,g/ml for times ranging from 1 hours to 24 hours. At lower doses (0 to .8uvg/mi) and shorter times (0 to 2 hours) disruption was incomplete and some remaining cortical rings were seen. Treatment with higher concentrations (> lug/ml) and longer times (> 2 hours) led to the complete loss of cortical actin rings. Under the latter conditions, the actin was relocalized as bright punctate accumulations throughout the cytoplasm of each cell. Rhodamine phallotoxin labeling indicated that the 42 actin cytoskeleton could be effectively and optimally disrupted by a treatment of lp:g/ml CD for 4 hrs (Fig 4vi). Phase microscopy of live cells indicated that the optimal treatment with CD caused cell rounding and retraction, and that it also caused spaces to form between the cells (Fig 5). Thus, like Ac treatment, CD treatment disrupted the integrity of the cell monolayer which suggests that cell-cell junctions were affected. 43 Fig 4: Scp2 mammary epithelial cell monolayers treated with disrupting agents (immunofluoresence) i) Vehicle control (undisrupted) monolayers stained for tubulin by indirect immunofluoresence. Microtubules form a filamentous network throughout the cytoplasm of each cell. ii) Microtubule cytoskeleton disrupted by treatment with 10 u.g/ml ND for 4hr. The microtubules are depolymerized and the tubulin is diffused throughout the cytoplasm. iii) Vehicle control (undisrupted) monolayers stained for cytokeratin by indirect immunofluoresence. Keratin filaments form a network throughout the cell cytoplasm that cage the nucleus. iv) Keratin cytoskeleton disrupted by treatment with 20mM Ac for 4 hr. The great majority of keratin filaments are disrupted and a significant amount of the staining is re-localized to the cell periphery. v) Vehicle control (undisrupted) monolayers stained for f-actin using rhodamine phallotoxin. Most of the filamentous actin is organized in cortical rings localized at cell-cell junctions. vi) Actin cytoskeleton disrupted by treatment with l|_ig/ml CD for 4 hr. F-actin re-localized to bright punctate accumulations throughout the cytoplasm. All the images were collected digitally using a confocal laser scanning microscope. 44 Fig 5: Morphological effects of cytoskeletal disruption on cell monolayers Vehicle control (undisrupted) cell monolayers displayed a classical 'cobblestone' epithelial morphology with no spaces between cells due to adherens junction formation When the monolayers were treated with lOug/ml ND for 4hours, they maintained a morphology similar to the control. Specifically, the cells remained flat and maintained cell-cell contacts. When cells were treated with 20mM Ac for 4 hours, the cells rounded and retracted slightly, cell-cell junctions were partially disrupted, and small spaces formed between the cells. When cells were treated with lu\g/ml CD for 4 hours, the cells rounded and retracted, cell-cell junctions were severely disrupted, and the spaces between cells were larger than that of Ac treated monolayers. The figure represents phase micrographs of live cultures. 46 Undisrupted N D HI Table 1: Optimal doses , t imes and effects of the disrupting agents on each individual cytoskeletal element and the cell morphology Dose Time Effect Nocodazole lOug/ml 4 hours - Complete disruption of microtubules - Staining diffused throughout the cytoplasm - Cells maintain cell-cell contacts Acrylamide 20 mM 4 hours - Disruption of keratin filaments - Increased staining at cell periphery - Disruption of cell-cell contacts and slight cell retraction Cytochalasin D lug/ml 4 hours - Complete disruption of actin filaments - Punctate staining throughout the cytoplasm - Disruption of cell-cell contacts and cell retraction 48 2. Disruption specificity All three cytoskeletal components can interact with each other through common linker proteins such as, plectin (Houseweart and Cleveland, 1998), BPAG1 (Yang et al., 1999), or the microtubule-associated protein tau (Henriquez et al., 1995). Therefore, it was important to identify the effects of the individual disrupting agents on all three major cytoskeletal components. To assay for disruption specificity, epithelial cells were plated as monolayers and were treated with the disrupting agents at the optimal conditions identified above. Cells were then fixed and were cross-stained for cytoskeletal elements that were not directly affected by each disrupting agent (data summarized in Table 2). 2.1. Nocodzole treatment Cell monolayers were treated with lOug/ml ND for 4 hours were fixed and stained for actin and keratin. In these monolayers, there was a decrease in the junction-associated actin cortical rings. Furthermore, the majority of filamentous actin was reorganized into stress fibers radiating throughout the cytoplasm. In contrast, the ND treatment had no effects on the keratin cytoskeleton. 2.2. Acrylamide treatment. Cell monolayers treated with 20 mM Ac for 4 hours were fixed and stained for actin and microtubules. The actin staining intensity was reduced and a punctate staining pattern was observed throughout the cytoplasm. The microtubule network, however, was unaffected by the Ac treatment. 49 2.3. Cytochalasin D treatment. Cell monolyers treated with 1 |Hg/ml CD for 4 hours were fixed and stained for microtubules and keratin. The staining pattern indicated that CD had no effect on microtubules. However, the keratin filaments were completely disrupted and the staining was reduced to very weak punctate spots. In summary, this cross-staining data indicates that both Ac and CD treatments were able to disrupt both keratin and actin cytoskeletons. In contrast, ND treatment had no effect On the keratin cytoskeleton while it caused a reorganization of much of the actin cytoskeleton into stress fibers. This lack of specificity for each disrupting agent will be commented upon in later sections. 50 Table 2: Effects of each disrupting agent on the microtubule, keratin and actin cytoskeletons Nocodazole Acrylamide Cytochalasin D Microtubules -Complete disruption -Diffuse staining -Undisrupted -Undisrupted Keratin -Undisrupted -Disrupted keratin filaments -Increased staining at cell periphery -Disrupted filaments -Weak, punctate staining Actin -Decreased cortical actin bundles -Reorganization into stress fibers. -Disrupted filaments -Punctate staining pattern -Reduced staining intensity -Disrupted filaments -Punctate staining pattern 51 3. Normal morphogenesis and differentiation of mouse mammary epithelial cells placed on a basement membrane 3.1 Morphogenesis When mammary epithelial cells are placed on a reconstituted basement membrane gel (Matrigel), they form polarized alveolus-like spheres with the basal surface of the cells interacting with the basement membrane and the apical surface of the cells facing the central lumen (Barcellos-Hoff et al., 1989). When scp2 mammary epithelial cells were placed on Matrigel, they attached and remained rounded. With time these cells aggregated, gathered matrix around them, and formed spherical structures 100-200 [im in diameter. Phase microscopy indicated that these spheres were tightly packed with sharp borders interacting with the Matrigel (Fig 6a). The adherens junction protein, E-cadherin, was localized at cell-cell contact sites throughout each sphere (Fig. 6b). The unstained region in the center of the sphere depicted the central lumen in these three-dimensional mammary epithelial clusters. In contrast to monolayers, the actin in three-dimensional cultures did not co-localize precisely with E-cadherin. Instead, the great majority of actin filaments was organized into bundles and formed a network at the apical region of each sphere surrounding the central lumen area. There were also a limited number of actin filaments radiating throughout each cluster (Fig. 6c). The tight junction protein occludin was localized at the most apical region of the cluster, surrounding the central lumen (Fig. 6d). Therefore, cells in the spheres were polarized. These sphere formation was quantified on the basis of size, shape, smooth edges of all the spheres on the culture surface. 52 These results indicate that scp2 mouse mammary epithelial cells placed on Matrigel were able to undergo complete morphogenesis forming polarized structures similar to the alveolus of the late pregnant mammary gland (These structures are identified as 'mammospheres' throughout this thesis). Given the requirement for dynamic changes in cell-cell and cell-ECM junctions during alveolar morphogenesis (Weaver et al, 1997; Lochter et al., 1997) it is likely that the cytoskeleton plays an important role in this process. 3.2 Differentiation Concomitant with alveolar morphogenesis, mammary epithelial cells on Matrigel functionally differentiate and express milk protein genes in the presence of lactogenic hormones (Roskelley et al., 1994). When scp2 cells were placed on Matrigel and allowed to undergo morphogenesis for three days in the presence of lactogenic hormones (insulin, prolactin, hydrocortisone), lactoferrin and P-casein gene expression was induced (Fig 7). In contrast, when the cells were placed on tissue culture plastic and allowed to grow as monolayers they did not induce either lactoferrin or P-casein. Matrigel-dependent induction of lactoferrin, which is regulated transcriptionally, is initiated by changes in cell shape (Close et al, 1997). P-Casein induction, which is also regulated transcriptionally, is initiated by the laminin-dependent activation of the a6P4 integrin (Streuli et al., 1995; Muschler et al., 1999). Cell shape and integrin signaling are facilitated, at least in part, by the cytoskeleton. Therefore, we expected that cytoskeletal disruption would also effect milk protein gene expression. 53 Fig 6: Basement membrane-dependent mouse mammary epithelial cell morphogenesis Scp2 mouse mammary epithelial cells were plated on Matrigel coated dishes and were allowed to undergo alveolar morphogenesis for 3 days, in serum-free media containing insulin, hydrocortisone and prolactin. Cells aggregate, gather matrix around them and formed single cell thickness spheres with a central lumen. Arrows indicate the apical domain facing the central lumen and arrow-heads indicate the basal domain of the mammosphere facing the Matrigel. a) Phase micrograph of a mammary epithelial mammosphere on Matrigel. Cells aggregated to form a tightly packed spherical cluster with a defined basal border that interacts with the matrix. b) A mammosphere, stained with an antibody against cell-cell adherens junction protein E-cadherin. E-cadherin was localized at cell-cell borders. The lack of staining in the center of the cluster defines the central lumen. c) A mammosphere, stained with rhodamine-labelled phallotoxin that binds to actin. A cortical actin network localized at the apical region surrounds the central lumen of the cluster. There were also a. few actin stress fibers radiating throughout the cluster toward the basal surface. d) A mammosphere stained with an antibody against the tight junction protein occludin. Occludin is only localized at the most apical region of the mammosphere at the edge of the central lumen. (b-d) Projected stacks of confocal laser scaning microscopic images of whole mamosphers 54 Fig 7: Mouse mammary epithelial cell differentiation on Matrigel Scp2 cells were cultured on tissue culture plastic (-) or on Matrigel-coated (+) dishes in the presence of lactogenic hormone for three days. Cells on plastic were scraped and collected while the cells on Matrigel were collected by Dispase treatment. 20 p:g of whole cell protein was separated by 13% SDS-PAGE and transferred on to PVDF membrane. The membrane was cut across the 50 kDa marker and the top half was probed with a polyclonal anti-lactoferrin antibody while the lower half was probed with a monoclonal anti-P-casein antibody. The cells cultured in the absence of Matrigel (on plastic) did not express either lactoferrin or P-casein. However, the cells that were cultured on Matrigel, expressed both lactoferrin (70kDa) and p-casein (30 kDa). P-casein degradation due to Dispase activity was seen as lower molecular weight bands below 30 kDa p-casein. (M = rainbow molecular weight markers) 56 Matrigel M - + 57 4. Effects of nocodazol treatment on the morphogenesis and differentiation of cells cultured on Matrigel 4.1 Effects on cell morphogenesis Mammary epithelial cells were plated on Matrigel-coated dishes and allowed to attach and aggregate. These aggregates were then treated with vehicle or increasing concentrations of ND for 4 hours and allowed to undergo morphogenesis for 72 hours. The cell aggregates that were treated with vehicle alone (control) formed large (~100pi in diameter), tightly packed mammospheres with sharp basal borders interacting with the Matrigel (Fig 8). Cells that were treated with 2u\g/ml ND formed mammospheres were morphologically similar to the control. In contrast, treatment with 9p:g/ml of ND disrupted mammosphere formation (Fig 8). These spheres were smaller in size and displayed a rough outer appearance due to loosely attached cells. There were also increased numbers of single cells relative to the 2p,g/ml condition. The treatments with 30p,g/ml and 100u\g/ml ND severely disrupted mammosphere formation. There were increasing numbers of single cells and smaller cell clusters. At 100p:g/ml ND the majority of cells remained single and those clusters that did form were ill defined. Immunofluoresence labelling for tubulin indicated that there was no difference in the microtubule network between the control cells and 2p.g/ml ND treated aggregates. They formed pattern of filamentous networks that radiated throughout to cell cytoplasm. However, the cell aggregates that were treated with 9u\g/ml, 30p,g/ml and 100p:g/ml showed a "bleached-out" diffused tubulin staining pattern indicating complete disruption of microtubules (data not shown). Thus, 9p.g/ml ND treatment was sufficient to disrupt 58 the microtubule network in cell aggregates and the cells were unable to undergo alveolar morphogenesis. This is consistent with the monolayer experiments where 10p:g/ml ND disrupted the microtubules. 4.2. Effects on cell differentiation Basement membrane-dependent lactoferrin and P-casein induction were both effectively inhibited by a 9p:g/ml ND treatment (Fig 9), which was also the minimal concentration required to disrupt microtubules and prevent alveolar morphogenesis. However, given the fact that lactoferrin induction, unlike P-casein, does not require morphogenesis it cannot be concluded from these experiments that the ability of ND to inhibit differentiation was restricted to its ability to disrupt mammosphere formation. 59 Fig 8: Cell cluster morphology of cel ls treated with nocodazole and recovered for 72 hours Scp2 cells were placed on Matrigel-coated dishes and were treated with the indicated doses of ND for 4 hours. ND was then removed and cells were allowed to recover (undergo morphogenesis) for 72 hours. Untreated control cells were organized into large nicely packed mammospheres with well-defined, smooth basal borders. The cells treated with 2u.g/ml ND underwent normal morphogenesis such that the morphology was similar to untreated cells. At 9 ug/ml and 30p:g/ml ND, cells manage to aggregate but were unable to organize into defined mammospheres. Cells that were treated with 100|J,g/ml ND were completely inhibited from undergoing morphogenesis, and mainly remained as single cells. All images, phase microscopy of live cell clusters. 60 Vehicle control Co 2/L/g/ml ND 9pg/ml ND 30 /^g/ml ND 100pg/ml ND Fig 9: Lactoferrin and (3-casein expression in cel ls cultured on Matrigel and treated with nocodazole i) Scp2 cells were allowed to attach to Matrigel overnight, treated with ND for 4 hour and recovered for 72 hour. Whole cell lysates (25|ig) were then separated by 13% PAGE and immunoblotted for the 70kDa lactoferrin and 30kDa 0-casein milk proteins. Both lactoferrin (70 kDa) and P-casein (30 kDa) protein expression were inhibited by treatment with ND concentrations greater than 2 ug/ml. A 75 kDa non-specific band was observed in all the samples (arrowhead, upper panel). ii) 12.5p.g of whole cell lysate protein for each treatment condition was separated by SDS-PAGE and was stained with coomassie-blue to demonstrate equivalent protein loading in each lane (M= rainbow markers). 62 5. Effects of acrylamide treatment on the morphogenesis and differentiation of cells cultured on Matrigel 5.1. Effects on cell morphogenesis and keratin filaments Cell aggregates were treated with vehicle or increasing concentrations Ac for 4 hours and then were allowed to undergo morphogenesis for 72 hours. Aggregates that were treated with lOmM Ac formed mammospheres that morphologically similar to the control. In contrast, treatment with 20mM Ac disrupted mammosphere formation (Fig 10). These spheres were smaller in size and they displayed a rough outer appearance due to loosely attached cells. There were also increased numbers of single cells relative to the lOmM condition. Treatment with 40mM Ac severely disrupted mammosphere formation. The majority of cells remained single and most of the clusters that did form were ill defined. When the cells were treated with 80mM Ac all the cells detached from the Matrigel (not shown). Immunofluoresence labeling for cytokeratin indicated that there was no difference in the keratin network between the control and lOmM Ac treated aggregates (data not shown). These aggregates formed filamentous networks radiating throughout to cell cytoplasm. However, with increasing concentration of Ac (20mM and 40mM), the number of visible keratin filaments was decreased and staining was more localized to the basal areas of each cell cluster, which showed a partially collapsed appearance in the keratin cytoskeleton. This was consistent with the monolayer experiments where 20mM Ac disrupted the keratin cytoskeleton. Thus, as was the case with ND, concentrations of Ac that disrupted the cytoskeleton prevented alveolar morphogenesis. 64 5.2. Effects on cell differentiation ECM dependent lactoferrin induction was unaffected by all Ac concentrations tested. However, P-casein induction was effectively inhibited at 20mM Ac treatment (Fig 11), which is the minimal concentration required to disrupt the keratin network and alveolar morphogenesis. Given the fact that lactoferrin induction, unlike P-casein, does not require morphogenesis these data suggest that the ability of Ac to inhibit differentiation was due to its ability to prevent mammosphere formation. The continued induction of lactoferrin also indicates that Ac treatment was not toxic. 65 Fig 10: Cell cluster morphology of cells treated with acrylamide and recovered for 72 hours Scp2 cells were allowed to attach to Matrigel overnight and were treated with the indicated doses of Ac for 4 hours. Ac was then removed and cells were allowed to recover (undergo morphogenesis) for 72 hours. Untreated vehicle controls (OmM Ac) organized into large nicely packed mammospheres with well defined smooth basal borders. The cells treated with 10 mM Ac underwent normal morphogenesis such that the morphology was similar to untreated cells. At 20mM Ac, cells managed to aggregate but the spheres were smaller and less well defined. Cells that were treated with 4()mM Ac were completely inhibited from undergoing morphogenesis, and mainly remained as single cells or ill-defined clumps of cells. All images, phase microscopy of live cell clusters. 66 Fig 11: Lactoferrin and p-casein expression in cells plated on Matrigel and treated with acrylamide. i) Scp2 cells were plated overnight, treated with Ac for 4 hr and recovered for 72 hr. Whole cell lysates (25ug) were then immunoblotted for the 70kDa lactoferrin and 30kDa P-casein milk proteins. Lactoferrin (70 kDa) protein expression was unaffected at all the tested Ac concentrations. However, p-casein (30 kDa) protein expression was partially inhibited by lOmM Ac treatment and fully inhibited by treatment by 20mM Ac treatment. ii) 12.5|ig of whole cell lysate protein for each treatment condition was separated by SDS-PAGE and was stained with coomassie-blue to demonstrate equivalent protein loading in each lane (M= rainbow markers). 68 6. Effects of cytochalasin D treatment on the morphogenesis and differentiation of cells cultured on Matrigel 6.1. Effects on cell morphogenesis Cell aggregates were treated with vehicle or increasing concentrations of CD for 4 hours, washed and allowed to undergo morphogenesis for 72 hours. Aggregates that were treated with .2 (ig/ml and .9(ig/ml CD underwent normal morphogenesis such that mammospheres formed were morphologically similar to the control. In contrast, treatment with lOug/ml CD subtly disrupted mammosphere formation. The latter spheres displayed a rough outer appearance due to loosely attached cells (Fig 12). Direct labeling for filamentous actin with rhodamine phallotoxin indicted that there were no differences in the actin filament network between the control aggregates and .2 p.g/ml and .9u\g/ml CD treated aggregates. The majority of actin was polarized and organized into bundled networks surrounding the apical region of each mammosphere. The apical actin networks were slightly disrupted in 3p.g/ml CD treatment. The aggregates that were treated with 10p;g/ml CD showed much reduced apical staining intensity as well as an increase in punctate cytoplasmic staining (Fig 13). Thus, 10p:g/ml CD treatment completely disrupted the actin network prevented proper alveolar morphogenesis. 6.2. Effects on cell differentiation ECM-dependent lactoferrin induction was unaffected by all CD concentrations tested. However, (3-casein induction was reduced at .9ug/ml and 3pig/ml 70 CD and completely inhibited at 10u.g/ml CD treatment (Fig 14). These CD concentrations had little effect on phase morpholigical organization as seen by phase microscopy. However 3u.g/ml CD did affect apical actin localization. This led me to speculate that cytoskeleton may have been affecting the ability of mammospheres to form actin-dependent cell-cell junctions. 71 Fig 12: Cell cluster morphology of cells treated with cytochalasin D and recovered for 72 hours Scp2 cells were placed on Matrigel-coated dishes and were treated with the indicated doses of CD for 4 hours. CD was then removed and cells were allowed to recover (undergo morphogenesis) for 72 hours. Untreated vehicle controls organized into nicely packed mammospheres with well defined smooth basal borders. The cells treated with < 3p:g/ml CD recovered completely and the morphology was similar to untreated cells. At 10p:g/ml CD, cells managed to form large aggregates but were unable to organize into defined mammospheres. These aggregates were disorganized and contained very rough basal edges with loosely-attached cells. All images, phase microscopy of live cell clusters. 72 .fy/g/ml CD 3/;g/ml CD 10/jg/mI CD 73 Fig 13: Act in localization of cell clusters treated with cytochalasin D and recovered for 72 hours Sep 2 cells were allowed to attach to Matrigel overnight and were treated with the indicated dose of CD for 4hours. CD was then removed and the cells were allowed to recover for 72 hours. Then the cells were fixed, stained for actin and images were digitally collected on a confocal laser scanning microscope. In the vehicle control actin bundles were localized to the apical region of the mammospheres. The actin filaments in the .2p:g/ml and .9|ig/ml CD treatment conditions were similar to the control. Treatment with 3(ig/ml CD slightly disrupted the apical actin network. In the aggregates that were treated with 10|ig/ml CD, there apical actin networks were absent and there was an increase in punctate cytoplasmic staining. 74 Fig 14: Lactoferrin and p-casein expression in cells cultured on Matrigel, treated with cytochalasin D i) Scp2 cells were plated overnight, treated with CD for 4 hours and recovered for 72 hours. Whole cell lysates (25p,g) were then immunoblotted for the 70kDa lactoferrin and 30kDa ^-casein milk proteins. Lactoferrin (70 kDa) protein expression was unaffected at all the tested Ac concentrations. However, p-casein (30 kDa) protein expression was reduced by treatment with .2pig/ml and 3p.g/ml CD and completely prevented by lOug/ml CD. ii) 12.5p.g of whole cell lysate protein for each treatment condition was separated by SDS-PAGE and was stained with coomassie-blue to demonstrate equivalent protein loading in each lane (M= rainbow markers). 76 17 7. Effects of cytochalasin D treatment on morphogenesis-dependent cell-cell junction formation 7.1 Cell-cell junctions of cell monolayers and mammosphers The actin cytoskeleton is associated with both E-cadherin containing adherens junctions and occludin containing tight junctions. Thus, I was interested in determining the role of these junctions in actin-dependent alveolar morphogenesis. Initially, to identify the junctions present in undifferentiated cells, I stained scp2 cell monolayers for actin, E-cadherin, and occludin by indirect immunofluoresence (Fig. 15). Both actin and E-cadherin were localized at cell-cell contacts sights forming cortical rings around the cells. However, epithelial monolayers were completely negative for occludin staining. When mammospheres on Matrigel were stained, E-cadherin was localized to cell-cell attachment sites throughout the mammosphere while occludin was localized specifically to the most apical border facing the central lumen as was the great majority of actin. These data indicated that E-cadherin mediated adherens junctions were formed in monolayers and thus, while they may be a prerequisite for morphogenesis, they do not actively contribute to the process. In contrast, occludin-mediated tight junctions were only formed when morphogenesis (ie. mammosphere formation) occurred. This led me to hypothesize that tight junction formation may be actively involved in the morphogenic process and the associated induction of p-casein expression. 78 7.2 The effect of cytochalasin treatment on tight junction formation Treatment of cells on Matrigel with 10|ig/ml of CD completely disrupted apical actin networks (Fig. 13). In other systems, the effects of CD are reversible. As such, I treated scp2 cell clusters on Matrigel with 10|ig/ml of the disrupting agent for 4 hours and then examined actin localization for a six-day recovery period (Fig. 16). For the first three days of recovery the actin was diffuse and punctate while on day 4 actin began to accumulate in the vicinity of the center of the clusters. This apical staining was more pronounced at day 5 of recovery and by day 6 it was indistinguishable from control, untreated (compare with Fig. 13 above). Thus the effects of CD treatment on actin localization were reversible in scp2 cell clusters on Matrigel. I next examined occludin localization to determine if 10p.g/ml CD would disrupt tight junctions, and, if this was the case, whether or not tight junctions would reform during the six day recovery period (Fig 17). For the first two days of the recovery period I did not detect any occludin localization (compare this with apical localization in controls; Fig. 15). By day 3 recovery there was a slight localization of occludin in the center of the clusters that became more pronounced throughout the remainder of the recovery period. Therefore, CD treatment did disrupted basement membrane-dependent tight junction formation and these junctions reformed with kinetics similar to that observed for apical actin network reformation after removal of the disrupting agent. Treatment of cell clusters on Matrigel with 10p.g/ml CD completely inhibited 0-casein induction (see Fig. 14). As a six day recovery period reversed the effects of this dose of CD on the actin cytoskeleton and tight junction formation I next examined (3-casein recovery during this period (Fig. 18). No P-casein induction was apparent in the 79 first three days of recovery when both actin and tight junctions were disrupted. In contrast, p-casein induction gradually increased through days 4-6 of recovery, the same period over which apical junctions re-formed. Therefore, morphogenesis-dependent differentiation (ie. p-casein expression) is strongly correlated with the initiation of apical/basal polarity. 1.3 Cell shape regulates tight junction formation When scp2 monolayrs were fixed and permeabilized, I could detect no occludin by immunoblotting (Fig. 15). However, the protein was present in monolayer cell lysates as assayed by Western blotting (Fig 19). Interestingly, all of the occludin was found in the soluble fraction rather than the cytoskeletal associated insoluble fraction. In contrast, the MDCK kidney epithelial line had considerable cytoskeletal-associated occludin in the western blot (Fig 19) and I could detect junctional occludin by immunostaining (data not shown). MDCK cells have tight junctions in the monolayer culture (Sakakibara et al., 1997) and were thus used as a positive control. Taken together, these data led me to conclude that cytoskeleton-associated tight junctional occludin is not present in the scp2 monolayers and that the absence of free cytoplasmic staining in these cells may be an artifact of my fixing and permeabilization procedure. One requirement of basement-membrane dependent mammary epithelial morphogenesis is cell rounding (Emerman and Pitelka, 1977; Streuli et al, 1990; Roskelley et al., 1994; Muschler et al., 1999). Therefore I was interested in determining if cell rounding alone would induce tight junction formation in scp2 cell clusters, even in the absence of the addition of basement membrane. To carry out such experiments I 80 coated tissue culture dishes with non-adhesive polymer polyHEMA (see methods for details) which allows the cells to attach to the substratum, but not spread. As a result cells form rounded clusters (Roskelley et al., 1995). Under these conditions a significant proportion of the occludin was relocalized to the cytoskeletal associated fraction (Fig 19). Some of this cytoskeletal-associated occludin underwent a slight upward mobility shift which may be indicative of tight-junction- dependent phosphorylation (Sakakibara et al., 1997). Unfortunately, I was unable to determine the subcellular localization of occludin by Western blotting in Matrigel cultures because of technical problems in extracting soluble and cytoskeletal-associated occludin under the latter conditions. It has been also shown that these rounded cells do not express P-casein in the absence of ECM proteins. 81 Fig 15: Actin, E-cadherin and occludin localization in mammary epithelial cell monolayers and mammospheres Scp2 mouse mammary epithelial cells were plated on tissue culture plastic or Matrigel-coated dishes and were allowed to undergo morphogenesis for 3 days. a. A monolayer stained with rhodamin-labled phallotoxin that binds to actin. Cortical actin network localized at the periphery of each cell where cell form contacts with other cells. b. A mammosphere, stained with rhodamine-labelled phallotoxin that binds to actin. A cortical actin network localized at the apical region surrounds the central lumen of the cluster. c. A monolayer stained with an antibody against cell-cell adherens junction protein E-cadherin. E-cadherin staining was localized to the same regions as the cortical actin rings in the monolayer. d. A mammosphere, stained with an antibody against cell-cell adherens junction protein E-cadherin. E-cadherin was localized at cell-cell borders. The lack of staining in the center of the cluster defines the central lumen. e. A monolayer stained with an antibody against tight junction protein occludin. The monolayer was completely negative for occludin. f. A mammosphere stained with an antibody against the tight junction protein occludin. Occludin is only localized at the most apical region of the mammosphere. Notice that actin was also localized to the same region. All the images are confocal laser scanning microscopic projections. 82 Fig 16: Actin cytoskeleton in cells plated on Matrigel, treated with 10(ig/ml cytochalasin D and allowed to recover for 6 days Cells were plated overnight on Matrigel-coated coverslips and were treated for 4 hours with 10u\g/ml CD. Samples of cells were fixed daily, stained for actin with rhodamine phallotoxin, and images were digitally collected on a confocal laser scanning microscope. At days 1 and 2, the actin cytoskeleton remained completely disrupted. By day 3 there were some actin bundles localized close to the apical region of the mammospheres. However, by day 4, the actin cytosleleton was completely recovered and majority of the actin was in bundle localized to apical regions caging the central lumen area of each mammosphere. 84 Day 1 Day 2 Fig 17: Occludin in cells plated on Matrigel, treated with 1 (hag/ml cytochalasin D and allowed to recover for 6 days Cells were plated overnight on Matrigel-coated coverslips and were treated with 10p:g/ml CD for 4 hours. Samples of cells were fixed daily, immuno-stained for occludin and images were digitally collected on a confocal laser scanning microscope. Occludin staining was completely negative in samples from days 1 and 2. There was some visible occludin staining at day 3. However, by days 4, 5 and 6 occludin was completely localized to the apical region surrounding the central lumen of each mammosphere. Arrows indicate the apical domain and arrow-heads indicate the basal domain of the mammosphere. 86 Fig 18: Lactoferrin and p-casein expression in cells plated on Matrigel, treated with CD and recovered for 6 days Scp2 cells were plated on Matrigel-coated tissue culture dishes and treated with 10p.g/ml CD for 4 hours. Cells were allowed to recover for 6 days and cell samples were collected every 24 hours. Whole cell lystes (20p:g) were then immunoblotted for lactoferrin and 0-casein. Lactoferrin protein expression was unaffected by CD treatment. However, (3-casein induction was restored at day 4 and reached maximum induction at day 6. 88 Fig 19: Occludin phosphorylation and membrane localization The soluble cytoplasmic fractions (S) and membrane fractions (M) of tight junction protein occludin was analyzed by Western blotting. MDCK cells were used as a control. Occludin is a 65 kDa protein and, depending on the phosphorylation stage, its molecular weight can range from 65 to 85 kDa. In MDCK cell monolayers, most of the occludin is in the membrane fraction, and higher MW bands up to 70 kDa was seen. A 65 kDa band was seen in the soluble fraction indicating the unphosphorylated occludin. When Scp2 cells monolayers were analyzed for occludin, all the occludin was located in the soluble fraction and there were none in the membrane fraction. The cells that were rounded-up and aggregated by plating them on polyHEMA coated dishes. In these cells, majority of occludin was localized to the membrane fraction and there were at molecular weights ranging from 65 kDa to 70 kDa indicating phosphorylation. The soluble fraction contained 65 kDa unphosphorylated occludin. 90 scp2 M D C K F L A T R O U N D E D s C s c s c 65 k D a ^ — » S=Soluble C--Cytoskeletal HI IV. DISCUSSION 1. Mammosphere formation and differentiation on reconstituted basement membrane (Matrigel) Unlike cells maintained as monolayers on tissue culture plastic, scp2 mammary epithelial cells placed on Matrigel rounded and aggregated to form clusters that became polarized and differentiated over period of days. Thus, these cell clusters, which closely resemble mammary alveoli in vivo, have been termed 'mammospheres' (Somasiri and Roskelley, 1999). Given the requirement for changes in cell shape, cell junction formation, and the initiation of cell polarity during mammosphere formation, I hypothesized that the cytoskeleton plays a role in this morphogenic process, and perhaps in the induction of tissue-specific differentiation (ie. milk protein gene expression) that accompanies it. Due to the paucity of systematic studies examining this subject in the literature, I began by disrupting each of the three major epithelial cytoskeletal components and determine the effects on mammosphere morphogenesis and differentiation. 2. Effects of nocodazole The primary action of ND is to depolymerize the microtubule cytoskeleton and release free tubulin (Rennison et al., 1992). When I treated scp2 cell monolayers with 10|ig/ml ND, microtubules were completely depolymerized both in monolayer and Matrigel cultures. While it did not affect monolayer morphology, ND treatment did have an interesting effect on the actin cytoskeleton. Specifically, it caused the breakdown of 92 cortical actin localized at cell-cell junctions and it increased the formation of cytoplasmic stress fibers which are normally not present in scp2 epithelial cells. This phenomenon has also been noted in serum starved swiss 3T3 cells (Bershadsky et al., 1996) and I suspect that it may weaken cell-cell junctions. This tentative conclusion is supported by the observation that ND treated scp2 cells were unable to efficiently cluster together when they were maintained on Matrigel. As a result, morphogenesis did not occur. During the course of this thesis I examined the basement membrane-dependent induction of two milk protein genes: lactoferrin, which is transcriptionally regulated by cell rounding only (ie. does not require morphogenesis; close et al, 1997); and p-casein, which is transcriptionally regulated by oc6p4 integrin-mediated morphogenic changes after contact with laminin in the basement membrane (Muschler et al., 1999). ND treatment completely prevented both lactoferrin and P-casein induction. This agrees with the previousley published observation that another microtubule disrupter, colchicine, inhibited milk protein synthesis in primary cultures of mammary epithelial cells (Blum and Wicha, 1988). However, because ND inhibited both lactoferrin and P-casein induction in the scp2 cells, I could not conclude that microtubule disruption specifically prevented milk protein gene expression because of its ability to prevent morphogenesis. At lower doses of ND (ie. lug/ml) Rennison et al. (1992) reported that milk protein secretion, but not milk protein synthesis, was inhibited in primary cultures. It is difficult to compare the continued synthesis of milk proteins that occurs in primary culture and the inductive synthesis that is being assayed in our scp2 cell line/Matrigel system. However, even though I did not examine secretion, I did find that similar low ND dose did not inhibit P-casein induction in the Matrigel cultures. Given that the induction 93 of P-casein by Matrigel is transcriptionally regulated (Schmidhauser et al., 1990; Roskelley and Bisseil, 1995), I concluded that the requirement for microtubules for milk protein synthesis and secretion are two separate differentiative events. 3. Effects of acrylamide In order to determine the role of intermediate filaments in mammary epithelial morphogenisis, I treated cells with the intermediate filament disrupting agent Ac. Ac is a neurotoxin that specifically alters the intermediate filament network by reversibly collapsing it (Durham et al, 1983; Eckert, 1985; Sager and Matheson, 1988). When scp2 cells were cultured as monolayers and treated with 20mM Ac, the keratin filament network collapsed and epithelial morphology was disrupted. The cells retracted slightly and small spaces formed between cells. Eckert and Yeagle, 1987 have made similar observations in PtKl cells and they suggest that the retraction of keratin filaments away from the cortical cytoplasm may be due to the effects of Ac on cell adhesion proteins at the membrane. While Ac, effects on the cell surface molecules were not specifically investigated in this study the inabihty of Ac treated cells to cluster on Matrigel are consistent with the conclusions of Eckert and Yeagle. The disruption of keratin filament in the monolayers also resulted in the disruption of the actin cytoskeleton. The keratin filaments could interact with the actin cytoskeleton through linker proteins such as BPAG1 (Yang et al, 1999) or they can interact directly. These interactions may be important for the stability of both cytoskeletal components. Therefore, Ac treatment also could be indirectly affecting the actin 94 associated cell-cell junctions thereby inhibiting alveolar morphogenesis and prehaps differentiation. In control cultures, cells formed well-defined mammospheres as described previously. Ac treatment leads to the disruption of this morphology. With the increasing concentration of Ac spheres became small and unorganized. There were increasing numbers of single cells and cells detached from the Matrigel suggesting diminished cell-cell, and cell-ECM adhesion and cell polarity. Keratin filaments are connected to structures known as hemidesmosomes by interacting with the oc6p4 integrin. These hemidesmosomes are required for the epithelial cell attachment to the basement membrane (Fuchs and Weber 1994). This was further confirmed by the observation that Ac treated scp2 cells detached from Matrigel and the extent of detachment increased with the Ac concentration. Hemidesmosomes are also important in initiating and defining the apical /basal polarity of the cells. Thus, detaching from the basement membrane also leads to the loss of cell polarity. Ac treatment also specifically inhibited the p-casein induction but not lactoferrin induction. This morphogenetic-specific interruption of differentiation further supports the requirement of oc6P4 integrin for P-casein gene induction. Contradictory to my observations, Seely and Aggeler (1991) demonstrate an increase in milk protein synthesis in primary mammary epithelial cells from mid pregnant mice. This discrepancy in observation could be due to the fact that my experiments were designed to induce morphogenesis while Seely and Aggeler maintained the cells that have undergone morpo genesis already. 95 At hemodesmosomes, bullous pemphigoid antigen 1 (BPAG1) links the keratin filaments to the a6P4 integrin. BPAG1 also contains an actin binding, thus can act as a linker protein between keratin filaments and actin filaments (Yang et al., 1996; Bernier et al., 1996; Leung et al, 1999). Therefore, Ac treatment could be indirectly responsible for disrupting the actin cytoskeleton and its associated cell-cell junctions. My findings from the Ac disruption studies lead to the conclusion that Ac treatment specifically disrupted the morphogenesis-dependent differentiation of mammary epithelial cells. 4. Effects of cytochalasin D To determine the role of filamentous actin in mammary epithelial morphogenesis I treated the cells with CD. The primary action of CD is to reversibly inhibit actin polymerization (Brown and Spudich., 1979; Cooper et al., 1987) and therefore, prevent filamentous actin network formation. When scp2 cell monolayers were treated with lp:g/ml CD the cortical actin network was completely disrupted as was cell morphology. The cells rounded and spaces formed between cells. The morphological effect was similar, although more exaggerated than that seen with Ac treatment, which further supports my tentative conclusion made above that the actin and intermediate filament cytoskeletons interact in these celLs. In control Matrigel cultures, actin bundles became localized at the apical end of scp2 cells such that they reorganize into filamentous network that formed a cage around the central lumen of the mammosphere. CD treatment disrupted this cage and the 96 morphology of the mammosphere was altered. While the cells did not completely dissociate from each other the spheres became ragged and loosely packed suggesting that cell-cell interactions and cell polarity were diniinished. Like Ac, CD treatment also specifically inhibited P-casein induction but not lactoferrin induction. Thus, this morphogenetic-specific interruption of differentiation was another piece of evidence that there is a linkage between the actin and intermediate cytoskeletons and that it is functionally important. My finding that CD treatment inhibited P-casein synthesis agrees with previously published data examining primary mammary epithelial cell cultures (Seeley and Aggeler et al., 1991) 5. Tight junction formation is correlated with p-casein induction Treatment with CD caused cell clusters on Matrigel to remain disorganized. This led me to further investigate the role of cell junctions in mammosphere morphogenesis. When scp2 cells were cultured as monolayers on tissue culture plastic they formed E-cadhem-containing adherens junctions but they did not form occludin-containing tight junctions. In contrast, when the cells were cultured as mammospheres on a basement membrane gel occludin became localized in apical cell-cell junction sites. Two observations indicated that this localization of occludin may be actin-dependent. Firstly, a great majority of the mammospheric actin co-localized with occludin. Secondly, high dose CD treatment completely disrupted the occludin staining. This is not a surprising finding given that tight junctions are intimately associated with the actin cytoskeleton. Interestingly, when CD was removed and the mammospheres were observed for an 97 extended recovery period actin was first to be relocalized to apical domains followed by occludin and another tight junction protein, ZO-1. Only then did P-casein induction occur, which suggests that tight junction formation may be one of the cellular changes initiated by the basement membrane that is required for milk protein gene induction. Given the function of the mammary gland, it is not unreasonable to suggest that the epithelial cells must form a permeability barrier before they can produce and secrete milk proteins. In polarized epithelial cells tight junctions contribute to just such a permeability barrier (Balda et al., 1996; Wong and Gumbiner, 1997; Sakakibara et al., 1997). In unpolarized scp2 mammary epithelial cell monolayers occludin was not localized to cell-cell junctions and it did not associate with the cytoskeleton. To determine if cell rounding, which is an important component of the mammary epithelial cell response to the basement membrane (Somasiri and Roskelley, 1999), alters occludin localization I used the non-adhesive substratum poly-HEMA in the absence of a basement membrane ECM. Under these conditions occludin relocalized to the membrane and it is associated with the cytoskeleton. In addition, the cytoskeletal occludin fraction underwent an upward mobility shift on Western blots, a phenomenon which has been implicated in the phosphorylation dependent localization of occludin to tight junctions in other systems (Sakakibara et al, 1997; Wong, 1997; Tsukamoto and Nigam, 1999). These data indicate that cell rounding may be sufficient to initiate tight junction formation. However, the junctions were not polarized in these 'naked' rounded cells. Furthermore, these unpolarized rounded cells do not express p-casein. When epithelial cells interact with laminin in the basement membrane, hemidesmosomes form. Hemidesmosomes act as basal anchors and they are required to 98 initiate apical/basal cell polarity across the epithelium (Fuchs and Weber, 1994). The ability of hemidesmosomes to initiate cell polarity is dependent on and is mediated, at least in part, by the interaction of the a604 integrin with the intermediate filament cytoskeleton (Giancotti, 1997). In mammary epithelial cells a6fi4 integrin interact with keratin intermediate filaments while, endothelial cell a6(34 integrin associates with vimentin filament network (Homan et al, 1998) to form hemidesomsome like structures. Recently it has been demonstrated that a6fj4 integrin activation is required for basement membrane-dependent P-casein expression by mammary epithelial cells (Muschler et al, 1999). Therefore, my findings that a disruption of the intermediate filament cytoskeleton prevents P-casein expression further supports the notion discussed above that the initiation of polarity is an important component of mammary epithelial cell morphogenesis and differentiation. The findings of the work outlined in this thesis can be summarized in a model indicating possible involvement of the cytoskeleton in mammary epithelial morphogenesis and differentiation (Fig. 20) 6. Conclusions. In this work I have demonstrated that morphogenesis and differentiation dependent P-casein gene expression in mouse mammary epithelial cells requires an intact cytoskeleton. The results of ND treatments were inconclusive. ND could have been toxic to the cells under these experimental conditions. However, given the many processes the microtubule network is involved it will be important to further investigate its role in the 99 mammary epithelial morphogenesis and differentiation. The results of the acrylmide studies indicate that the keratin cytoskeleton was also important for the proper differentiation of the mammary epithelial cells, as indicated by the specific inhibition of P-casein expression. Further, my data agrees that a6P4 integrin is a major signaling receptor in P-casein protein induction. Therefore, specific signaling processes driven through ct6P4 integrin receptor should be identified in further studies. Adding Matrigel to pre-rounded and clustered cells in the presence of a6P4 blocking antibodies will provide information with respect the role of a6P4 integrin the actual process of morphogenesis. Actin disruption studies showed that the actin cytoskeleton was important in initial polarity and cell-cell junction dynamics. Given the dynamic nature of the mammary gland it is important to regulate gene expression with the state of differentiation. In this study I demonstrate that epithelial cells of mammosphers in culture had to completely polarize and form tight junctions at the apical region in the process of morphogenesis before they induced milk protein P-casein. Also the cells undergoing morphogenesis, reorganized their cytoskeleton accordingly to accommodate the changes to the cell-cell junctions. Thus, the effects of the oc6P4 integrin in the induction of the P-casein'gene may be mediated through cell-cell interacting junction formation. This is important for the proper physiological function of the gland, which is to secrete, contain and transport milk proteins to the nipple of the gland. 100 Fig. 20: Summary model for possib le role of the cytoskeleton in mammary epithelial morphogenesis and differentiation Actin Keratin Microtubules Adherens Junctions (E-cadherin) Tight junctions_ (occludin) Hemidesmosome (a6(34 - laminin) Apical/basal polarity Morphogenesis Differentiation (ie. P-casein) a6p4 integrin - Gene expression - Secretion 101 V. REFERENCES Allan VJ, and RD Vale. 1991. Cell cycle control of microtubule-based membrane transport and tubule formation in vitro. J Cell Biol. 113(2):347-59. Alberts B, Bray D, Lweis J, Raff M, Roberts K and J Watson. 1994. The cytoskeleton. In Molecular biology of the cell. Garland Publishing, Inc. 803-815. Aggeler J, Ward J, Blackie L M , Barcellos-Hoff MH, Streuli CH, and MJ Bissell. 1991. Cytodifferentiation of mouse mammary epithelial cells cultured on a reconstituted basement membrane reveals striking similarities to development in vivo. J. Cell Sci. 99 ( Pt2):407-17. Ando-Akatsuka Y, Saitou M, Hirase T, Kishi M, Sakakibara A, Itoh M , Yonemura S, Furuse M , and S Tsukita. 1996. Interspecies diversity of the occludin sequence: cDNA cloning of human, mouse, dog, and rat-kangaroo homologues. J Cell Biol. 133(l):43-7. Balda MS, Whitney JA, Flores C, Gonzalez S, Cereijido M, and K Matter. 1996. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical-basolateral intramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J Cell Biol. 134(4): 1031-49. Barcellos-Hoff MH, Aggeler J, Ram TG, and MJ Bissell. 1989. Functional differentiation and alveolar morphogenesis of primary mammary cultures on reconstituted basement membrane. Development. 105(2):223-35. Bernier G, Mathieu M, De Repentigny Y, Vidal SM, and R Kothary. 1996. Cloning and characterization of mouse ACF7, a novel member of the dystonin subfamily of actin binding proteins. Genomics. 38(1): 19-29. Bershadsky A, Chausovsky A, Becker E, Lyubimova A, and B Geiger. 1996. Involvement of microtubules in the control of adhesion-dependent signal transduction. Curr Biol. 6(10):1279-89. Best A, Ahmed S, Kozma R, and L Lim. 1996. The Ras-related GTPase Racl binds tubulin. J Biol Chem. 271(7):3756-62. Blum JL, and MS Wicha. 1988. MS Role of the cytoskeleton in laminin induced mammary gene expression. J Cell Physiol. 135(1): 13-22. Brown SS and JA Spudich. 1979. Cytochalasins inhibits the rate of elongation of actin filament fragments. J. Cell Biol., 83:657-662. Buxton RS, Co win P, Franke WW, Garrod DR, Green KJ, King I A, Koch PJ, Magee A l , 102 Rees DA, and JR Stanley. 1993. Nomenclature of the desmosomal cadherins. J Cell Biol. 121(3):481-3. Review. Citi S, Sabanay H, Jakes R, Geiger B, J Kendrick-Jones. 1988. Cingulin, a new peripheral component of tight junctions. Nature. 333(6170):272-6. Citi S. 1993. The molecular organization of tight junctions.J Cell Biol. 121(3):485-9. Review. Citi S, Volberg T, Bershadsky AD, Denisenko N, and B Geiger. 1994. Cytoskeletal involvement in the modulation of cell-cell junctions by the protein kinase inhibitor H-7. J Cell Sci. 107 ( Pt 3):683-92. Chen H, Paradies NE, Fedor-Chaiken M, R Brackenbury. 1997. E-cadherin mediates adhesion and suppresses cell motility via distinct mechanisms. J Cell Sci. 110 ( Pt 3):345-56. Clark EA, King WG, Brugge JS, Symons M and RO Hynes. 1998. Integrin-mediated signals regulated by members of the rho family of GTPases. J Cell Biol. 27;142(2):573-86. Close MJ, Howlett AR, Roskelley CD, Desprez PY, Bailey N, Rowning B, Teng CT, Stampfer MR, and P Yaswen. 1997. Lactoferrin expression in mammary epithelial cells is mediated by changes in cell shape and actin cytoskeleton. J. Cell Sci. 110 ( Pt 22):2861-71. Cooper JA. 1987. Effects of cytochalasin and phalloidin on actin. J Cell Biol. 105(4): 1473-8. Review. Crowe DT, Chiu H, Fong S, and IL Weissman. 1994. Regulation of the avidity of integrin alpha 4 beta 7 by the beta 7 cytoplasmic domain. J Biol Chem. 269(20): 14411-8. Daniel CW and GB Silberstein. 1987. "Postnatal development of the rodent mammary gland" In The mammary gland: development, regulation and function. Plenum press, New York. 3-36. Danielson KG, Oborn CJ, Durban EM, Butel JS, and D Medina. 1984. Epithelial mouse mammary cell line exhibiting normal morphogenesis in vivo and functional differentiation in vitro. Proc Natl Acad Sci U S A . 81(12):3756-60. Dedhar S, and GE Hannigan. 1996. Integrin cytoplasmic interactions and bidirectional transmembrane signalling. Curr Opin Cell Biol. 8(5):657-69. Review. Desprez PY, Roskelley CD, Campisi J and MJ Bisseil. 1993. Isolation of functional cell lines from a mouse mammary epithelil cell strain: the importance of basement membrane and cell-cell interactions. Mol. Cell Differentiation. 1:99-110. 103 Dowling J, Yu QC, and E Fuchs. 1996. Beta4 integrin is required for hemidesmosome formation, cell adhesion and cell survival. J Cell Biol. 134(2):559-72. Durham HD, Pena SD, S Carpenter. 1983. The neurotoxins 2,5-hexanedione and acrylamide promote aggregation of intermediate filaments in cultured fibroblasts. Muscle Nerve. 6(9):631-7. Eckert BS, and PL Yeale. 1987. NMR investigation of acrylamide-induced changes in the intermediate filament organization. J. Cell Biol., 105:422a. Eckert BS. 1985. Alteration of intermediate filament distribution in PtKl cells by acrylamide. Eur J Cell Biol. 37:169-74. Emerman, J. T., and D. R. Pitelka. 1977. Maintenance and induction of morphological differentiation in dissociated mammary epithelium on floating collagen membranes. In Vitro, 13(5):316-328. Emerman, J. T., Burwen, S. J., and D. R. Pitelka. 1979. Substrate properties influencing ultrastructural differentiation of mammary epithelial cells in culture. Tissue and cell, 11(1):109-119. Emerman, J. T., Enami, J., Pitelka, D. R., and S. Nandi. 1977. Hormonal effects on intracellular and secreted casein in cultures of mouse mammary epithelial cells on floating collagen membranes. Proc. Natl. Acad. Sci. USA, 74(10): 4466-4470. Frixen UH, Behrens J, Sachs M , Eberle G, Voss B, Warda A, Lochner D, and W Birchmeier. 1991. E-cadherin-mediated cell-cell adhesion prevents invasiveness of human carcinoma cells. J Cell Biol. 113(l):173-85. Fuchs E, and K Weber. 1994. Intermediate filaments: structure, dynamics, function, and disease. Annu Rev Biochem. 63:345-82. Review. Furuse M, Fujimoto K, Sato N, Hirase T, Tsukita S, S Tsukita. 1996. Overexpression of occludin, a tight junction-associated integral membrane protein, induces the formation of intracellular multilamellar bodies bearing tight junction-like structures. J Cell Sci. 109 ( Pt 2):429-35. Furuse M, Fujita K, Hiiragi T, Fujimoto K, and S Tsukita. 1998. Claudin-1 and -2: novel integral membrane proteins localizing at tight junctions with no sequence similarity to occludin. J Cell Biol. 141(7): 1539-50. Furuse M, Hirase T, Itoh M, Nagafuchi A, Yonemura S, Tsukita S, aand S Tsukita. 1993. Occludin: a novel integral membrane protein localizing at tight junctions. J Cell Biol. 123(6 Pt 2): 1777-88. 104 Furuse M, Itoh M, Hirase T, Nagafuchi A, Yonemura S, Tsukita S, and S Tsukita. 1994. Direct association of occludin with ZO-1 and its possible involvement in the localization of occludin at tight junctions. J Cell Biol. 127(6 Pt l):1617-26. Giancotti FG. 1997. Integrin signaling: specificity and control of cell survival and cell cycle progression. Curr Opin Cell Biol. 9(5):691-700. Review. GaUicano GI, Kouklis P, Bauer C, Yin M, Vasioukhin V, Degenstein L, and E Fuchs. 1998. Desmoplakin is required early in development for assembly of desmosomes and cytoskeletal linkage. J Cell Biol. 143(7):2009-22. Gorska, A. E., Joseph, H., Derynck, R., Moses, H. L. and R Serra. 1998. Dominant-negative interference of the transforming growth factor beta type II receptor in mammary gland epithelium results in alveolar hyperplasia and differentiation in virgin mice. Cell Growth Differ. 9(3):229-38. Guo L, Degenstein L, Dowling J, Yu QC, Wollmann R, Perman B, and E Fuchs. 1995. Gene targeting of BPAG1: abnormalities in mechanical strength and cell migration in stratified epithelia and neurologic degeneration. Cell. 81(2):233-43. Green KJ, Goldman RD, and RL Chisholm. 1988. Isolation of cDNAs encoding desmosomal plaque proteins: evidence that bovine desmoplakins I and II are derived from two mRNAs and a single gene. Proc Natl Acad Sci U S A . 85(8):2613-7. Gumbiner BM. 1996. Cell adhesion: The molecular basis of tissue architecture and morphogenesis. Cell, 84:345-357. Gumbiner B. 1987. Structure, biochemistry, and assembly of epithelial tight junctions. Am J Physiol. 253(6 Pt l):C749-58. Review. Gumbiner BM, nd PD McCrea. 1993. Catenins as mediators of the cytoplasmic functions of cadherins. J Cell Sci Suppl. 17:155-8. Review. Henriquez JP, Cross D, Vial C, RB Maccioni. 1995. Subpopulations of tau interact with microtubules and actin filaments in various cell types.Cell Biochem Funct. 13(4):239-50. Hieda Y, Nishizawa Y, Uematsu J, and K Owaribe. 1992. Identification of a new hemidesmosomal protein, HD1: a major, high molecular mass component of isolated hemidesmosomes. J Cell Biol. 116(6): 1497-506 Haskins J, Gu L, Wittchen ES, Hibbard J, BR Stevenson. 1998. ZO-3, a novel member of the MAGUK protein family found at the tight junction, interacts with ZO-1 and occludin. J Cell Biol. 141(1): 199-208. 105 Hogervorst F, Kuikman I, von dem Borne AE, and A Sonnenberg. 1990. Cloning and sequence analysis of beta-4 cDNA: an integrin subunit that contains a unique 118 kd cytoplasmic domain. EMBO J. 9(3):765-70. Homan SM, Mercurio AM, and SE LaFlamme. 1998. Endothelial cells assemble two distinct alpha6beta4-containing vimentin-associated structures: roles for ligand binding and the beta4 cytoplasmic tail J Cell Sci. I l l ( Pt 18):2717-28. Hotta K, Tanaka K, Mino A, Kohno H, and Y Takai. 1996. Interaction of the Rho family small G proteins with kinectin, an anchoring protein of kinesin motor. Biochem Biophys Res Commun. 225(l):69-74. Ffouseweart MK, and DW Cleveland. 1998. Intermediate filaments and their associated proteins: multiple dynamic personalities. Curr Opin Cell Biol. 10(1):93-101. Review. Howlett AR, and MJ Bissell. 1993. The influence of tissue microenvironment (stroma and extracellular matrix) on the development and function of mammary epithelium. Epithelial Cell Biol. 2(2):79-89. Review. Howlett AR, Bailey N, Damsky C, Petersen OW, and MJ Bissell. 1995. Cellular growth and survival are mediated by beta 1 integrins in normal human breast epithelium but not in breast carcinoma. J Cell Sci. 108 ( Pt 5): 1945-57. Hynes RO. 1992. Integrins: versatility, modulation, and signaling in cell adhesion. Cell. 1992 69(1): 11-25. Review. Ingber DE, Prusty D, Sun Z, Betensky H, and N Wang. 1995. Cell shape, cytoskeletal mechanics, and cell cycle control in angiogenesis. J Biomech. 28(12):1471-84. Itoh M, Nagafuchi A, Moroi S, and S Tsukita. 1997. Involvement of ZO-1 in cadherin-based cell adhesion through its direct binding to alpha catenin and actin filaments. J Cell Biol. 138(l):181-92. Itoh M, Nagafuchi A, Yonemura S, Kitani-Yasuda T, Tsukita S, and S Tsukita. 1993. The 220-kD protein colocalizing with cadherins in non-epithelial cells is identical to ZO-1, a tight junction-associated protein in epithelial cells: cDNA cloning and immunoelectron microscopy. J Cell Biol. 121(3):491-502. Joseph, H., Gorska, A. E., Sohn, P., Moses, H. L. and R. Serra. 1999. Overexpression of a kinase-deficient transforming growth factor-beta type II receptor in mouse mammary stroma results in increased epithelial branching. Mol. Biol. Cell 10(4): 1221-34. Kelly DE, and A M Kuda. 1981. Traversing filaments in desmosomal and hemidesmosomal attachments: freeze-fracture approaches toward their characterization. AnatRec. 199(1):1-14. 106 Kleinman HK, McGarvey ML, Hassell JR, Star VL, Cannon FB, Laurie GW and GR Martin. 1986. Basement membrane complexes with biological activity. Biochemistry. 25(2):312-8. Knudson CM, Stemberger BH, and S Patton. 1978. Effects of colchicine on ultrastructure of the lactating mammary cell: membrane involvement and stress on the Golgi apparatus. Cell Tissue Res. 195(1):169-81. Krane, I. M. and P. Leder. 1996. NDF/heregulin induces persistence of terminal end buds and adenocarcinomas in the mammary glands of transgenic mice. Oncogene 12(8):1781-1788. Koch PJ, Walsh MJ, Schmelz M, Goldschmidt MD, Zimbelmann R, and WW Franke. 1990. Identification of desmoglein, a constitutive desmosomal glycoprotein, as a member of the cadherin family of cell adhesion molecules. Eur J Cell Biol. 53(1): 1-12. Koukoulis GK, Virtanen I, Korhonen M, Laitinen L, Quaranta V, and VE Gould. 1991. Immunohistochemical localization of integrins in the normal, hyperplastic, and neoplastic breast. Correlations with their functions as receptors and cell adhesion molecules. Am. J. Pathol. 139(4):787-99. Kouklis PD, Hutton E, and E Fuchs. 1994. Making a connection: direct binding between keratin intermediate filaments and desmosomal proteins. J Cell Biol. 127(4): 1049-60. Kratochwil, K. 1969. Organ specificity in mesenchymal induction demonstrated in the embryonic development of the mammary gland of the mouse. Dev. Biol. 20: 46-71. Kreis TE, Matteoni R, Hollinshead M, and J Tooze. 1989. Secretory granules and endosomes show saltatory movement biased to the anterograde and retrograde directions, respectively, along microtubules in AtT20 cells. Eur J Cell Biol. 49(1): 128-39. Kurpakus MA, and JC Jones. 1991. A novel hemidesmosomal plaque component: tissue distribution and incorporation into assembling hemidesmosomes in an in vitro model. Exp Cell Res. 194(1): 139-46. Lee EY, Parry G, and MJ Bisseil. 1984. Modulation of secreted proteins of mouse ammary epithelial cells by the collagenous substrata. J. Cell Biol. 98(1): 146-55. Leung CL, Sun D, and RK Liem.1999. The intermediate filament protein peripherin is the specific interaction partner of mouse BPAGl-n (dystonin) in neurons. J Cell Biol. 144(3):435-46. Li ML, Aggeler J, Farson DA, Hatier C, Hassell J, and MJ Bisseil. 1987. Influence of a reconstituted basement membrane and its components on casein gene expression and secretion in mouse mammary epithelial cells. Proc Natl Acad Sci U S A . 84(1): 136-40. 107 Lin CQ, Dempsey PJ, Coffey RJ, and MJ Bissell. 1995. Extracellular matrix regulates whey acidic protein gene expression by suppression of TGF-alpha in mouse mammary epithelial cells: studies in culture and in transgenic mice. J Cell Biol.l29(4):l 115-26. Lochter A, Galosy S, Muschler J, Freedman N, Werb Z, and MJ Bissell. 1997. Matrix metalloproteinase stromelysin-1 triggers a cascade of molecular alterations that leads to stable epithelial-to-mesenchymal conversion and a premalignant phenotype in mammary epithelial cells. J Cell Biol. 139(7): 1861-72. Lochter A, Galosy S, Muschler J, Freedman N, Werb Z, and MJ Bissell. 1997. Matrix metalloproteinase stromelysin-1 triggers a cascade of molecular alterations that leads to stable epithelial-to-mesenchymal conversion and a premalignant phenotype in mammary epithelial cells. J Cell Biol. 139(7): 1861-72. Martin G, Rohrabach H, Terranova V and L Loitta. 1982. Function and pathology of basement membranes. Ch. 3 In: Connective tissue diseases. Raven press, N.Y., 16-31. Morita K, Furuse M, Fujimoto K, and S Tsukita. 1999. Claudin multigene family encoding four-transmembrane domain protein components of tight junction strands. Proc Natl Acad Sci U S A . 96(2):511-6. Munemitsu S, Souza B, Muller O, Albert I, Rubinfeld B, and P Polakis. 1994. The APC gene product associates with microtubules in vivo and promotes their assembly in vitro. Cancer Res. 54(14):3676-81. Muschler J, Lochter A, Roskelley CD, Yurchenco P nd MJ Bissell. 1999. Division of labor among the oc6(34 integrin, (31 integrins, and an E3 laminin receptor to signal morphogenesis and P-casein expression in mammary epithelial cells. Mol Biol Cell. 10(9):2817-28. Myers CA, Schmidhauser C, Mellentin-Michelotti J, Fragoso G, Roskelley CD, Casperson G, Mossi R, Pujuguet P, Hager G, and MJ Bissell. 1998. Characterization of BCE-1, a transcriptional enhancer regulated by prolactin and extracellular matrix and modulated by the state of histone acetylation. Mol Cell Biol. 18(4):2184-95. Nelson WJ, Shore EM, Wang AZ, and RW Hammerton. 1997. Identification of a membrane-cytoskeletal complex containing the cell adhesion molecule uvomorulin (E-cadherin), ankyrin, and fodrin in Madin-Darby canine kidney epithelial cells. J Cell Biol. 110(2):349-57. Niemann, C , Brinkmann, V., Spitzer, E., Hartmann, G., Sachs M. , Naundorf, H. and W. Birchmeier. 1998. Reconstitution of mammary gland development in vitro: requirement of c-met and c-erbB2 signaling for branching and alveolar morphogenesis. J. Cell Biol. 143(2):533-545. 108 Nobes CD, and A Hall. 1995a. Rho, rac and cdc42 GTPases: regulators of actin structures, cell adhesion and motility. Biochem Soc Trans. 23(3):456-9. Review. Nobes CD, and A Hall. 1995b. Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. CeU. 81(l):53-62. Parry G, Lee EY, Farson D, Koval M, and MJ Bisseil. 1985. Collagenous substrata regulate the nature and distribution of glycosaminoglycans produced by differentiated cultures of mouse mammary epithelial cells. Exp Cell Res. 156(2):487-99. Paulsson M, Saladin K, and E. Engvall. 1991. Structure of laminin variants. The 300-kDa chains of murine and bovine heart laminin are related to the human placenta merosin heavy chain and replace the a chain in some laminin variants. J Biol Chem. 266(26):17545-51. Pitelka, D. R., Taggart, B. N., and S. T. Hamamoto. 1983. Effects of extracellular calcium depletion on membrane topography and occludin junctions of mammary epithelial cells in culture. J. Cell Biol. 96: 613-624. Pignatefli M, Hanby AM, and GW Stamp. 1991. Low expression of beta 1, alpha 2 and alpha 3 subunits of VLA integrins in malignant mammary tumours. J. Pathol. 165(1):25-32. Rajasekaran AK, Hojo M, Huima T, and E Rodriguez-Boulan. 1996. Catenins and zonula occludens-1 form a complex during early stages in the assembly of tight junctions. J Cell Biol. 132(3):451-63. Rennison ME, Handel SE, Wilde CJ, and RD Burgoyne. 1992. Investigation of the role of microtubules in protein secretion from lactating mouse mammary epithelial cells. J Cell Sci. 102 ( Pt 2):239-47. Rasbridge SA, Gillett CE, Sampson SA, Walsh FS, and RR Millis. 1993. Epithelial (E-) and placental (P-) cadherin cell adhesion molecule expression in breast carcinoma. J Pathol. 169(2):245-50. Roskelley CD, and MJ Bisseil. 1995. Dynamic reciprocity revisited: a continuous, bidirectional flow of information between cells and the extracellular matrix regulates mammary epithelial cell function. Biochem Cell Biol. 73(7-8):391-7. Roskelley CD, Srebrow A, and MJ Bisseil. 1995. A hierarchy of ECM-mediated signalling regulates tissue-specific gene expression. Curr Opin Cell Biol. 7(5):736-47. Review. 109 Roskelley CD, Desprez PY, aand MJ Bissell. 1994. Extracellular matrix-dependent tissue-specific gene expression in mammary epithelial cells requires both physical and biochemical signal transduction. Proc Natl Acad Sci U S A . 20;91(26): 12378-82. Sager PR, and DW Matheson. 1988. Mechanisms of neurotoxicity related to selective disruption of microtubules and intermediate filaments. Toxicology. 49(2-3):479-92. Sakakibara A, Furuse M, Saitou M, Ando-Akatsuka Y, and S Tsukita. 1997. Possible involvement of phosphorylation of occludin in tight junction formation. J Cell Biol. 137(6):1393-401. Sakakura, T. 1987. Mammary embryogenesis in "Mammary gland development, regulation, and function". Plenum press, New york. 37-63. Sawamura D, Li K, Chu ML, and J Uitto. 1991. Human bullous pemphigoid antigen (BPAG1). Amino acid sequences deduced from cloned cDNAs predict biologically important peptide segments and protein domains. J Biol Chem. 266(27): 17784-90. Schmidhauser C, Bissell MJ, Myers CA, Casperson GF. 1990. Extracellular matrix and hormones transcriptionally regulate bovine beta-casein 5' sequences in stably transfected mouse mammary cells. Proc Natl Acad Sci U S A . 87(23):9118-22. Seely KA, and J Aggeler. 1991. Modulation of milk protein synthesis through alteration of the cytoskeleton in mouse mammary epithelial cells cultured on a reconstituted basement membrane. J Cell Physiol. 146(1): 117-30. Shienvold FL, and DE . Kelly. 1976. The hemidesmosome: new fine structural features revealed by freeze-fracture techniques. Cell Tissue Res. 172(3):289-307. Shimizu Y, Van Seventer GA, Horgan KJ, and S Shaw. 1990. Regulated expression and binding of three V L A (beta 1) integrin receptors on T cells. Nature. 345(6272):250-3. Silberstein GB, and CW Daniel. 1982. Glycosaminoglycans in the basal lamina and extracellular matrix of the developing mouse mammary duct. Dev Biol. 90(l):215-222. Simpson KJ, Bird P, Shaw D, and K Nicholas. 1998. Molecular characterisation and hormone-dependent expression of the porcine whey acidic protein gene. J Mol Endocrinol. 20(l):27-35. Simons K, and SD Fuller. 1985. Cell surface polarity in epithelia. Annu Rev Cell Biol. 1:243-88. Review. Shiozaki H, Tahara H, Oka H, Miyata M, Kobayashi K, Tamura S, Iihara K, Doki Y, Hirano S, and M Takeichi. 1991. Expression of immunoreactive E-cadherin adhesion molecules in human cancers. Am J Pathol. 139(l):17-23. 110 Small V, Rotter K and I Kaverina. 1999. Functional design in the actin cytoskeleton. Curr. Opin. Cell Biol. 11:54-60 Somasiri A and CD Roskelley. 1999. Cell shape and integrin signaling regulate the differentiation state of mammary epithelial cells. Methods in Molecular Biology, Integrin protocols, Humana press, 129:271-283 Somasiri A, Wu c, Ellchuk T and CD Roskelley. 1999. Phosphatidylinositol 3-kinase is required for adherens junction-dependent mammary epithelial morphogenesis. Submitted to Differentiation. Spinardi L, Einheber S, Cullen T, Milner TA, and FG Giancotti. 1995. A recombinant tail-less integrin beta 4 subunit disrupts hemidesmosomes, but does not suppress alpha 6 beta 4-mediated cell adhesion to laminins. J Cell Biol. 129(2):473-87. Spinardi L, Ren YL, Sanders R, FG Giancotti. 1993. The beta 4 subunit cytoplasmic domain mediates the interaction of alpha 6 beta 4 integrin with the cytoskeleton of hemidesmosomes. Mol Biol Cell. 4(9):871-84. Sternlicht MD, Lochter A, Sympson CJ, Huey B, Rougier JP, Gray JW, Pinkel D, Bisseil MJ, and Z Werb. 1999. The stromal proteinase MMP3/stromelysin-1 promotes mammary carcinogenesis. Cell. 98(2): 137-46. Stevenson BR, Anderson JM, and S Bulhvant. 1988. The epithelial tight junction: structure, function and preliminary biochemical characterization. Mol Cell Biochem. 83(2): 129-45. Review. Stevenson BR, Siliciano JD, Mooseker MS, DA Goodenough. 1986. Identification of ZO-1: a high molecular weight polypeptide associated with the tight junction (zonula occludens) in a variety of epithelia. J Cell Biol. 103(3):755-66. Streuli CH, Schmidhauser C, Bailey N, Yurchenco P, Skubitz AP, Roskelley C, Bisseil MJ. 1995. Laminin mediates tissue-specific gene expression in mammary epithelia. J. Cell Biol. 129(3):591-603. Streuli CH. and MJ Bisseil. 1990. Expression of extracellular matrix components is regulated by substratum. J. Cell Biol. 110: 1405-1415. Streuli, C. H , Schmidhauser, C , Kobrin, M. , Bisseil, M. J. and R Derynck. 1993. Extracellular matrix regulates expression of the TGF-beta 1 gene. J. Cell Biol. 120(l):253-60. Suzuki S, and Y Naitoh. 1990. Amino acid sequence of a novel integrin beta 4 subunit and primary expression of the mRNA in epithelial cells. EMBO J. 9(3):757-63. I l l Sympson, C. J., Talhouk, R. S., Alexander, C. M. , Chin, J. R., Clift, S.M., Bissell M . J and Z. Werb. 1994. Targeted expression of stromelysin-1 in mammary gland provides evidence for a role of proteinases in branching morphogenesis and the requirement for an intact basement membrane for tissue-specific gene expression. J. Cell Biol. 125(3):681-93. Tamura RN, Rozzo C, Starr L, Chambers J, Reichardt LF, Cooper HM, and V Quaranta. 1990. Epithelial integrin alpha 6 beta 4: complete primary structure of alpha 6 and variant forms of beta 4. J Cell Biol. 111(4): 1593-604. Teng, C. T., Pentecost, B. T., Chen, Y. H., Newbold, R. R., Eddy, E. M and J. A. McLachlan. 1989. Lactotransferrin gene expression in the mouse uterus and mammary gland. Endocrinology. 124(2):992-999. Topper, Y. J. and C. S. Freeman. 1980. Multiple hormone interactions in the developmental biology of the mammary gland. Physiol. Rev. 60(4): 1049-106. Tooze J, and B Burke. 1987. Accumulation of adrenocorticotropin secretory granules in the midbody of telophase AtT20 cells: evidence that secretory granules move anterogradely along microtubules. J Cell Biol. 104(4): 1047-57. Tsukamoto T, and SK Nigam. 1999. Role of tyrosine phosphorylation in the reassembly of occludin and other tight junction proteins. Am J Physiol. 276(5 Pt 2):F737-50. van der Neut R, Krimpenfort P, Calafat J, Niessen CM, and A Sonnenberg. 1996. Epithelial detachment due to absence of hemidesmosomes in integrin beta 4 null mice. Nat Genet 13(3):366-9. Warburton, M. J., Mitchell, D., Ormerod, E. J., and P. Rudland. 1982. Distribution of myoepithelial cells and basement membrane proteins in the resting, pregnant, lactating and involuting rat mammary gland. J. Histochem. Cytochem. 30: 667-676. Warburton, M. J., Monoghan, P., Ferns,S. A., Rudland, P. S., Perusinghe, N., and A. E. Chung. 1984. Distribution of entactin in the basement membrane of the rat mammary gland. Expl cell Res. 152: 240-254. Weaver V M , Petersen OW, Wang F, Larabell CA, Briand P, Damsky C, and MJ Bissell. 1997. Reversion of the malignant phenotype of human breast cells in three-dimensional culture and in vivo by integrin blocking antibodies. J Cell Biol. 137(l):231-45. Wong V, and BA Gumbiner. 1997. A synthetic peptide corresponding to the extracellular domain of occludin perturbs the tight junction permeability barrier. J Cell Biol. 136(2):399-409. Wong V. 1997. Phosphorylation of occludin correlates with occludin localization and function at the tight junction. Am J Physiol. 273(6 Pt l):C1859-67. 112 Yang Y, Bauer C, Strasser G, Wollman R, Julien IP, and E Fuchs. 1999. Integrators of the cytoskeleton that stabilize microtubules. Cell 98(2):229-38. Yang Y, Dowling J, Yu QC, Kouklis P, Cleveland DW, and E Fuchs. 1996. An essential cytoskeletal linker protein connecting actin microfilaments to intermediate filaments. Cell. 23;86(4):655-65. Yang, Y., Spitzer, E., Meyer, D., Sche, M. , Niemann, C , Hrtmann, G., Weidner, K. M. , Birchmeier, C and W. Birchmeier. 1995. Sequential requirement of hepatocyte growth factor and neuregulin in the morphogenesis and differentiation of the mammary gland. J. Cell Biol. 131(1): 215-226. Yant J, Buluwela L, Niranjan B, Gusterson B, and T Kamalati. In vivo effects of hepatocyte growth factor/scatter factor on mouse mammary gland development. Exp Cell Res. 241(2):476-81. 113 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089197/manifest

Comment

Related Items