Author's Accepted Manuscript

Scrib is Required for Epithelial Cell Identityand Prevents Epithelial To MesenchymalTransition in the Mouse

Idella F. Yamben, Rivka A. Rachel, ShaliniShatadal, Neal G. Copeland, Nancy A. Jenkins,Soren Warming, Anne E. Griep

PII: S0012-1606(13)00513-7DOI: http://dx.doi.org/10.1016/j.ydbio.2013.09.027Reference: YDBIO6224

To appear in: Developmental Biology

Received date: 15 April 2013Revised date: 3 September 2013Accepted date: 23 September 2013

Cite this article as: Idella F. Yamben, Rivka A. Rachel, Shalini Shatadal, Neal G.Copeland, Nancy A. Jenkins, Soren Warming, Anne E. Griep, Scrib is Requiredfor Epithelial Cell Identity and Prevents Epithelial To Mesenchymal Transitionin the Mouse, Developmental Biology, http://dx.doi.org/10.1016/j.ydbio.2013.09.027

This is a PDF file of an unedited manuscript that has been accepted forpublication. As a service to our customers we are providing this early version ofthe manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting galley proof before it is published in its final citable form.Please note that during the production process errors may be discovered whichcould affect the content, and all legal disclaimers that apply to the journalpertain.

www.elsevier.com/locate/develop-

mentalbiology

Scrib�������������� ���������������������������������� ����������

��������������������������������������

Idella F. Yambena, Rivka A. Rachelb1, Shalini Shatadala, Neal G. Copelandb2, Nancy A. Jenkinsb2, Soren Warmingb3, Anne E. Griepa*I

aDepartment of Cell and Regenerative Biology, School of Medicine and Public Health, University of Wisconsin-Madison, Madison, WI 53706 and

bMouse Cancer Genetics Program, Center for Cancer Research, National Cancer Institute, Frederick, MD 21702

*Corresponding author. Tel.: +6082 62 8988; fax: 6082 62 7306.

aegriep@wisc.edu

Abstract

The integrity and function of epithelial tissues depends on the establishment and maintenance of

defining characteristics of epithelial cells, cell-cell adhesion and cell polarity. Disruption of these

characteristics can lead to the loss of epithelial identity through a process called epithelial to

mesenchymal transition (EMT), which can contribute to pathological conditions such as tissue fibrosis

and invasive cancer. In invertebrates, the epithelial polarity gene scrib plays a critical role in

establishing and maintaining cell adhesion and polarity. In this study we asked if the mouse homolog,

Scrib, is required for establishment and/or maintenance of epithelial identity in vivo. To do so, we

conditionally deleted Scrib in the head ectoderm tissue that gives rise to both the ocular lens and the

corneal epithelium. Deletion of Scrib in the lens resulted in a change in epithelial cell shape from

cuboidal to flattened and elongated. Early in the process, the cell adhesion protein, E-cadherin, and

apical polarity protein, ZO-1, were downregulated and the myofibroblast protein, �SMA, was

upregulated, suggesting EMT was occurring in the Scrib deficient lenses. Correlating temporally with

the upregulation of �SMA, Smad3 and Smad4, TGF� signaling intermediates, accumulated in the

nucleus and Snail, a TGF� target and transcriptional repressor of the gene encoding E-cadherin, was

������������������������������������������������������������1 Present address: Neurobiology Neurodegeneration and Repair Laboratory, National Eye Institute, Bethesda, MD 20892. 2 Present address: The Methodist Cancer Research Program, The Methodist Academy, Houston, TX 77030. 3 Present address: Molecular Biology, Genentech, San Francisco, CA 94080.

� 2

upregulated. Pax6, a lens epithelial transcription factor required to maintain lens epithelial cell identity

also was downregulated. Loss of Scrib in the corneal epithelium also led to molecular changes

consistent with EMT, suggesting that the effect of Scrib deficiency was not unique to the lens. Together,

these data indicate that mammalian Scrib is required to maintain epithelial identity and that loss of Scrib

can culminate in EMT, mediated, at least in part, through TGF� signaling.

�

Highlights�

�

� Conditional�deletion�of�Scrib�in�the�lens�leads�to�epithelial�to�mesenchymal�transition.�� The�first�molecular�change�observed�was�disruption�of�E�cadherin�followed�by�loss�of�

ZO�1�and�Pax6.�� Upregulation�of�the�mesenchymal�protein��SMA�coincided�with�downregulation�of�

Pax6.�� Activation�of�TGF��signaling�coincided�with�upregulation�of��SMA.�� Conditional�deletion�of�Scrib�in�the�corneal�epithelium�leads�to�epithelial�to�

mesemchymal�transition.�

Keywords

Scrib,�PDZ�proteins,�Lens�development,�Epithelial�to�mesenchymal�transition,�Cell�adhesion,�Cell�polarity��

INTRODUCTION

The integrity and function of epithelial tissues depends on the establishment and maintenance of

defining properties of epithelial cells, cell-cell adhesion and cell polarity, which are composed of

cell junctional complexes including tight and adherens junctions. Junctional complexes,

furthermore, are essential factors in organogenesis through their linkages to the cytoskeleton and

the associated signaling pathways that modulate cell migration, proliferation, differentiation and

� 3

death (Baum and Georgiou, 2011; Giepmans and van Ijzendoorn, 2009). Identifying the factors

that are required for the establishment and maintenance of epithelial identity is crucial because of

their roles in normal development and homeostasis and because loss of epithelial identity is

associated with disease processes such as fibrosis and cancer (Acloque et al., 2009; Kalluri and

Weinberg, 2009; Thiery et al., 2009). In Drosophila, the PDZ domain containing protein

Scribble has been shown to play a role in establishing and maintaining epithelial cell adhesion,

polarity and proliferation (Bilder, 2004; Bilder et al., 2000; Bilder and Perrimon, 2000). Recent

studies suggest that the mouse homolog, Scrib, may perform similar and additional functions in

mammalian species (Montcouquiol et al., 2003; Murdoch et al., 2003; Pearson et al., 2011; Qin

et al., 2005). In this study, we address the possibility that Scrib is required for epithelial identity

in vivo in the mouse.

The failure to maintain epithelial cell properties can result in a change in cell identity

from an epithelial cell to a mesenchymal cell through a process referred to as epithelial to

mesenchymal transition (EMT). EMT is an important normal biological process contributing to

embryonic development (Nakaya and Sheng, 2008), wound healing (Kalluri and Weinberg,

2009), and tissue repair (Kalluri and Weinberg, 2009). Aberrant activation of EMT can

contribute to diseased states such as tissue fibrosis in the kidney, liver, lung, and ocular lens

(Kalluri and Weinberg, 2009), and to malignant progression of tumors (Acloque et al., 2009;

Humbert et al., 2008; Klymkowsky and Savagner, 2009; Saika et al., 2008; Tsuji et al., 2009).

EMT is characterized by the disassembly of adhesion junctions, loss of apical-basal polarity, and

the acquisition of migratory capacity (Kalluri and Weinberg, 2009; Tsuji et al., 2009). A number

of signaling pathways, including TGF�, are involved in promoting EMT during both

development and disease (Kalluri and Weinberg, 2009; Thiery et al., 2009). The phenotypic

� 4

changes found in EMT are associated with molecular changes including reduced expression of

epithelial proteins such as the adherens junction protein, E-cadherin, and the tight junction

protein, zonula occludens 1 (ZO-1), and upregulation in the expression of mesenchymal proteins

(Kalluri and Weinberg, 2009; Thiery et al., 2009), and proteins involved in remodeling of the

extracellular matrix such as certain matrix metalloproteinases (Dwivedi et al., 2006).

While it is known that factors such as E-cadherin and ZO-1 are essential components of

cell adhesion and polarity, and their loss promotes EMT, much less is known about the upstream

factors that are required to establish and maintain epithelial structure in vertebrates. In

invertebrates such as Drosophila, the PDZ (PSD-95, Dlg-1, and ZO-1) domain containing

protein, Scribble (Scrib), has been shown to be necessary for the establishment, positioning and

maintenance of adherens junctions and apical determinants (Bilder et al., 2000; Bilder and

Perrimon, 2000; Woods et al., 1996). Scrib is a leucine repeat and PDZ (LAP) protein containing

16 leucine rich-repeats and four PDZ domains (Johnson and Wodarz, 2003). The PDZ domain is

a protein-protein interaction domain of 80-90 amino acids found in common amongst members

of this family of proteins which assemble large macromolecular complexes involved in polarity,

vesicle transport and signaling in various cell types (Harris and Lim, 2001; Humbert et al.,

2003). In Drosophila, scrib mutants exhibit loss of apical-basal polarity that correlates with

hyperproliferation and loss of tissue architecture. scrib has been shown to act as a neoplastic

tumor suppressor (Bilder, 2004; Bilder et al., 2000; Bilder and Perrimon, 2000) and scrib-

deficient follicular cells exhibit mislocalization of basolateral junction proteins such as E-

cadherin (Zhao et al., 2008). Scrib-depleted MDCK cells were less adherent apparently due to

compromised E-cadherin function (Qin et al., 2005). Analysis of the mouse mutant, Crc, which

expresses a truncated form of Scrib in which the last two PDZ domains are absent (Murdoch et

� 5

al., 2003), has shown that Scrib is essential for planar cell polarity (Montcouquiol et al., 2003),

for maintaining epithelial cohesion during lung development (Yates et al., 2013) and for

angiogenesis (Michaelis et al., 2013). Finally, loss of Scrib in conjunction with expression of an

oncogenic K-ras gene in mice promotes prostate carcinogenesis (Pearson et al., 2011). Although

these findings indicate roles for Scrib in epithelial adhesion and polarity, whether and through

what mechanism Scrib is required specifically for establishing or maintaining the epithelial

identity in vivo in mammals has yet to be elucidated.

Previously, we reported that Scrib is expressed in the mouse ocular lens (Nguyen et al.,

2005). The lens is an ideal in vivo model for studying the mechanisms of establishing and

maintaining epithelial identity during embryogenesis. It is composed entirely of epithelial cells,

facilitating biochemical analysis, and the tissue itself is dispensable for the animal’s viability,

allowing one to follow the effects of loss of gene function throughout the life of the animal.

Development of the mouse lens begins around embryonic day 9.5 (E9.5) when a discrete region

of the head ectoderm is induced to thicken and form the lens placode, which then invaginates

along with the optic cup, the future retina (Piatigorsky, 1981). The invaginating lens detaches

from the overlying head ectoderm, forming the lens vesicle. Cells in the vesicle opposite the

optic cup differentiate into lens epithelial cells while cells closest to the optic cup differentiate

into lens fibers. By E13.5, the anterior surface of the lens is composed of a monolayered

epithelium overlaying a mass of elongated cells, the primary fiber cells, that arise from

differentiation of the cells in the posterior of the lens vesicle. From that point, the lens grows in

size as the self-renewing epithelium gives rise to cells, which, as they undergo terminal

differentiation, are added to the fiber cell compartment.

� 6

EMT is associated with two types of lens cataract, posterior capsular opacification (PCO)

and anterior subcapsular cataract (ASC). Using murine model systems, some aspects of the

mechanism through which these cataracts develop have been identified (de Iongh et al., 2005;

Saika et al., 2008; Saika et al., 2009; Wormstone et al., 2009). For example, conditionally

inactivating the gene encoding E-cadherin early in lens development (Pontoriero et al., 2009) or

the polarity protein aPKC� (Sugiyama et al., 2009) results in EMT. Also, overexpression of

self-activating form of TGF� in transgenic mouse lens (Srinivasan et al., 1998) and treatment of

rat lenses ex vivo (Hales et al., 1995) and lens explants in culture with TGF�2 result in

cataractous plaques expressing markers of EMT (Symonds et al., 2006). Importantly, in humans,

PCO is also associated with TGF� driven EMT as treatment of human capsular bags with

TGF�2 results in matrix wrinkling and expression of EMT markers (Wormstone et al., 2002).

Herein, we assessed the impact on epithelial cell identity of conditionally inactivating

Scrib in the precursor cells of the embryonic surface ectoderm that give rise to the corneal

epithelium and the lens. We found that inactivating Scrib resulted in lens defects that were

associated with the progressive downregulation of epithelial proteins and upregulation of the

mesenchymal protein, �SMA. Furthermore, we identified TGF� as a potential signaling

pathway mediating the EMT phenotype that arises in the lens as a consequence of loss of Scrib.

Finally, loss of Scrib also led to changes consistent with EMT in the corneal epithelium,

suggesting together with previous literature that Scrib is required to maintain epithelial identity,

and the integrity of epithelial tissues, in a variety of epithelia in vivo.

MATERIALS AND METHODS

� 7

Animal Maintenance and Use

All experiments using mice conformed to the Public Health Service Policy on Humane

and Use of Laboratory Animals and ARVO statement for the Use of Animals in Ophthalmic and

Vision Research and were approved by the Institutional Animal Care and Use Committee of the

University of Wisconsin-Madison School of Medicine and Public Health.

Generation of Scrib Conditional Null and Scrib Germline Null Mice To conditionally delete Scrib prior to the vesicle stage, Lens-Cre (LeCre) transgenic mice

(Ashery-Padan et al., 2000) were mated to Scribf/+ mice. Progeny containing the floxed allele

were identified using ScribF and Scrib1-X2-R primers, producing a wild type band of 257 bp and

floxed allele product of 325 bp (Table S1). Progeny containing the Cre recombinase gene were

identified using Cre-1 and Cre-3 primers, producing a single band of 420 bp (Table S1). Next,

Scribfl/+;LeCre mice were mated to Scribfl/+ mice to produce litters containing Scribfl/fl;LeCre

mice. The LeCre mice were maintained in the hemizygous on the FVB/N genetic background.

No overt ocular abnormalities were observed in these mice. Mice carrying a germline mutation

in Scrib were generated by crossing Scribfl/+ mice to ACTB-Cre mice (Lewandoski et al., 1997)

and Scrib+/- progeny were intercrossed to produce Scrib-/- day E17.5 embryos.

To confirm Cre-mediated deletion in the lens, genomic DNA samples isolated from P2

microdissected lenses from Scribfl/fl and Scribfl/fl;LeCre, and the eye and tail from E17.5 Scrib-/-

embryos were subjected to PCR amplification using the ScribF/ScribwtR and ScribF/ScribckoR

primer pairs (Table S1).

� 8

Histological Analysis

Embryos from E11.5, E13.5, heads from E15.5 and eyes from E17.5 and P10 Scribfl/fl and

ScribLecre mice were dissected, fixed overnight in 4% paraformaldehyde (PFA) at 4°C and

embedded in paraffin. Sections (5 �m) were cut, rehydrated, and stained with hematoxylin and

eosin and viewed by light microscopy as previously described (Rivera et al., 2009). In addition,

to verify that expression of the LeCre transgene itself did not lead to lens abnormalities, eyes

from stock LeCre mice and Scrib+/+;LeCre mice (generated via the intercrosses between

Scribfl/+;LeCre and Scribfl/+ mice) were fixed, embedded, sectioned and stained with hematoxylin

and eosin and viewed by light microscopy. No lens abnormalities were observed. Embryos were

staged by designating midday on the day of the vaginal plug as embryonic day 0.5. Postnatal

animals were staged by designating the day of birth as neonate (neo) and subsequent days as

postnatal day 1 (P1), P2, etc.

Immunofluorescence

Paraffin embedded sections were prepared as described above. To detect Pax6, E-

cadherin, �-catenin, ZO-1, and Smad4, and Scrib, rehydrated sections were boiled in sodium

citrate buffer (0.01 M, pH 6.0) in a rice cooker for 30 minutes, cooled, washed in 1X PBS,

blocked as previously described (Rivera et al., 2009), and incubated with primary antibody

overnight at 40°C (see Table S2 for antibody sources and dilutions). Following incubation with

primary antibody, sections were washed 1X PBS then incubated with FITC-conjugated horse

anti-mouse (Vector Laboratories), AlexaFluor 568 conjugated goat anti-rabbit, or AlexaFluor

568 conjugated donkey anti-goat (Molecular Probes) antibodies for one hour at room

� 9

temperature (RT) then washed and viewed by confocal microscopy. For Smad4 staining, nuclei

were counterstained with To-Pro 3 (Invitrogen) and for Pax6 and ZO-1 staining, nuclei were

counterstained with propidium iodide.

To detect Snail or Smad3, rehydrated sections were treated for 10 min in H2O2 (0.3%) in

methanol to quench endogenous peroxidases prior to antigen retrieval. Sections were blocked for

one hour in 5% rabbit serum diluted in 1X PBS at RT. The sections were incubated with SNA1

primary or Smad3 antibodies (Table S2) diluted in 2% (or 5% for Smad3) blocking solution

overnight at 4°C (SNA1) or one hour at RT (Smad3). Samples were washed in PBS and treated

with an anti-goat biotinylated secondary antibody for one hour at 1:100 at RT (or 1:200 for

Smad3) diluted in 2% blocking solution. Positive cells were identified using a Goat Elite

Vectastain ABC Kit (Vector Laboratories) and DAB peroxidase substrate kit (Vector

Laboratories). Nuclei were counterstained in hematoxylin. Sections were viewed by light

microscopy. Smad3 and Snail positive nuclei were counted on 6-8 sections from at least 3

different E17.5 Scribfl/fl and ScribLeCre eyes and the data subjected to statistical analysis using

the two-sided Wilcoxon Rank Sum test and p<0.05 was considered to be significantly different.

To immunostain for �SMA, cryosections were prepared from E17.5 and P10 Scribfl/fl and

ScribLeCre eyes, as previously described (Rivera et al., 2009). Sections (10 �m) were incubated

in �SMA-Cy3 conjugated antibody overnight at 4°C (Table S2) washed, counterstained with To-

Pro 3, and viewed by confocal microscopy. To immunostain for cytokeratin 12, cryosections

were incubated with primary antibody overnight at 4°C, washed and incubated with AlexaFluor

568 conjugated donkey anti-goat antibody (Molecular Probes) for one hour at RT, washed,

counterstained with To-Pro3, and viewed by confocal microscopy. To immunostain the cornea

for E-cadherin, cryosections from P10 Scribfl/fl and ScribLeCre were incubated with primary

� 10

antibody overnight at 40°C, washed, incubated with FITC conjugated rabbit anti-rat antibody

(Vector Laboratories) for one hour at RT, washed, and counterstained with propidium iodide.

Western Blotting

Protein lysates of P2 whole lenses and P10 corneas from Scribfl/fl and ScribLeCre mice and

protein lysates of brain tissue from E17.5 Scrib-/- embryos were prepared in 1X RIPA buffer with

protease inhibitors, as described previously (Rivera et al., 2009). A total of 50-100 μg of each

lysate was run on a 7.5% acrylamide gel, the proteins transferred to polyvinylidene difluoride

(PVDF) membranes, and the membranes incubated with anti-Scrib, anti-Pax6, and anti-E-

cadherin antibodies (Table S3). Membranes were washed in 1X PBST, incubated with goat anti-

mouse horseradish peroxidase (HRP, Pierce) or donkey anti-goat HRP (Santa Cruz) diluted in

blocking solution. After incubation with secondary antibodies, blots were washed and bands

visualized using Enhanced Chemiluminescence Plus Kit (ECL, Plus, GE Healthcare) and

exposed to film or scanned on a StormScanner phosphorimager. Blots were stripped and

reprobed with anti-Gapdh as a loading control. Bands were quantified and subjected to statistical

analysis using the two-sided Wilcoxon Rank Sum test and p<0.05 was considered to be

significantly different.

RESULTS

Generation of Scrib Conditional Mutants To study the role of Scrib in lens development, we generated a mouse strain carrying a

conditional allele of Scrib. In brief, a targeting vector with loxP sites flanking exons 2-8 and frt

sites flanking a positive neomycin (neo) marker was introduced into the mouse genome (Figure

� 11

1A). Mice transmitting this allele were mated with transgenic mice expressing Flippase in the

germline to remove the neo marker thereby generating the conditional null allele we used in this

study. These mice were mated to LeCre transgenic mice, in which the Cre recombinase enzyme

is expressed beginning at day E9.0 in the surface ectoderm overlying the optic vesicle and is

continually expressed in surface ectoderm-derived ocular structures such as the lens and corneal

epithelium (Ashery-Padan et al., 2000). This Cre line was chosen because Scrib expression is

observed in the invaginating lens pit at day E10.5 where it localizes primarily along the apical

surface of the lens vesicle, and to a lesser extent along the lateral membranes (Figure 1B). The

tissue specificity and timing of Cre expression in the ScribLeCre mice were confirmed by

monitoring expression of a Cre-activatable GFP reporter that is incorporated into the LeCre

transgene (Figure 1C).

To demonstrate that Cre-mediated deletion had occurred in the lens, genomic DNA was

prepared for PCR analysis from lenses of P2 Scribfl/fl and Scribfl/fl;LeCre mice (hereafter referred

to as control and ScribLeCre mice, respectively), eye tissue from day E17.5 embryos Scrib-/-, and

tails from Scribfl/fl, Scrib+/,and Scribfl/+mice and Scrib-/- embryos. Cre-mediated deletion of

exons 2-8 was evident in DNA from the lens of ScribLeCre and eye of Scrib-/-, but not in lens

DNA from control mice (Figure 1D). A faint band corresponding to the floxed allele was

observed in the ScribLeCre lens DNA. Protein lysates were prepared from the lenses of control

and ScribLeCre mice as well as lysates from Scrib+/+ and Scrib-/- brain. Western blot analysis

was performed using an anti-Scrib antibody and membranes were reblotted for Gapdh as a

loading control. Scrib levels in ScribLeCre lens lysates were reduced compared to the control,

although not completely absent (Figure 1E), consistent with the PCR results (Figure 1D). Scrib

was not detected in the Scrib-/- brain protein lysate, indicating the specificity of the antibody.

� 12

Scrib was detected in the corneal lysates from control mice but not in the corneal lysates from

ScribLeCre mice (Figure 1E). Together, these results indicate that Cre-mediated deletion of Scrib

in the lens and cornea was efficient, although possibly not complete in the case of the lens.

Loss of Scrib Results in Lens and Cornea Defects

Conditional loss of Scrib resulted in numerous ocular defects. By postnatal day 10 (P10)

ScribLeCre eyes (Figure 2A) and the dissected ScribLeCre lenses (Figure 2B) were notably

smaller than the controls; lenses were also misshapen and had central opacity, indicating the

presence of a cataract (Figure 2B, arrow). Hematoxylin and eosin staining of lens sections

showed that the ScribLeCre eyes had vacuolated lenses (Figure 2D, arrow/arrowhead) and a

hyperplastic iris, which occluded the pupil (Figure 2D, asterisk). High magnification images of

the lens epithelium revealed further defects. The control lens epithelium was a monolayer of

cuboidal cells, whereas the ScribLeCre lens epithelium was unrecognizable as a lens epithelium

and consisted of cells that appeared squamous rather than cuboidal (Figure 2E, F). The defects in

the epithelial and/or fiber cells could account for the observed cataracts and small eye size. High

magnification images of the corneal epithelium revealed defects. The control corneal epithelium

was stratified with prominent cuboidal cells in the basal layer (Figure 2G). However, the

ScribLeCre corneal epithelium had irregularly arranged, squamous cells (Figure 2H). These

defects in corneal and lens epithelial shape in the ScribLeCre mice suggested a possible loss of

epithelial identity.

Lens Epithelial Cells Lose Epithelial Characteristics and Acquire Mesenchymal Traits

� 13

To assess the possibility that the lens epithelial cells were undergoing a cell fate change

reflective of their abnormal morphology, we characterized the molecular nature of these cells.

Lens sections from P10 mice were subjected to immunofluorescence assays for the lens

epithelial proteins E-cadherin and, Pax6, a transcription factor required for expression of lens

epithelial specific crystallin genes (Cvekl et al., 2004). Robust, nuclear staining for Pax6 (green)

was observed in the epithelial cells of the control lenses where it overlapped with the nuclear

counterstain, propidium iodide (PI, red) (Figure 3A, yellow). However in the ScribLeCre

epithelial cells, some nuclei appeared red rather than yellow (Figure 3C, arrows) and in

unmerged images, these same nuclei appeared to lack Pax6 staining (Figure 3D, arrows). Other

nuclei were orange (Figure 3C) and showed reduced levels of Pax6 staining (Figure 3D).

Western blot analysis of extracts from P2 control and ScribLeCre lenses and quantification of the

results confirmed that Pax6 levels were reduced in ScribLeCre lenses by 55% (Figure 4).

Immunoflourescence staining for E-cadherin (green) and �-catenin (red) showed that these

proteins colocalized along all membranes of control epithelial cells (Figure 3E, yellow), and

unmerged image showed consistent E-cadherin staining along the membranes (Figure 3F). In

contrast, E-cadherin and �-catenin staining on ScribLeCre lenses showed regions in the

epithelium where staining was absent or reduced (Figure 3G, arrows) which was also evident in

the unmerged E-cadherin image (Figure 3H, arrows). Western blot analysis of extracts from P2

lenses and quantification of the results confirmed that E-cadherin levels were reduced by 35%

(p<0.04) in the ScribLeCre lenses at that time point (Figure 4).

The reduced expression of epithelial markers in the lens epithelium of the ScribLeCre

mice prompted us to investigate if these cells were undergoing EMT. �SMA is a myofibroblast

protein that commonly appears during EMT in the lens (de Iongh et al., 2005). Therefore,

� 14

cryosections from control and ScribLeCre mice were immunostained with anti-�SMA antibodies

(red) and counterstained with To-Pro3 (blue) to identify nuclei (Figure 3I-L). While �SMA was

present only in the iris in control lenses (Figure 3I, J), in ScribLeCre eyes staining for �SMA was

observed in the lens as well as in the hyperplastic iris (Figure 3K, L). Thus, some cells in the

lens epithelium of the ScribLeCre mice appear to have lost epithelial identity and acquired

mesenchymal identity.

EMT progression can be linked to activation of signaling through members of the TGF�

superfamily. Normally, the lens epithelium does not respond to TGF� signaling; however, EMT

in the lens is driven by TGF� (de Iongh et al., 2005; Wormstone et al., 2009). To determine if

signaling through the TGF� superfamily was now active in lens epithelial compartment, sections

from control and ScribLeCre mice were immunostained for Smad4, a component of the Smad

complex that accumulates in the nucleus during signaling through TGF� superfamily members

(Itoh et al., 2000). In control lenses, Smad4 was cytoplasmic in epithelial cells (Figure 3M, N).

However, Smad4 was concentrated in the nuclei of epithelial cells in the ScribLeCre lenses

(Figure 3O, P). Thus, cells in the ScribLeCre lens showed evidence for active signaling via the

TGF� superfamily.

Loss of E-Cadherin in ScribLeCre Lenses Begins Early in Lens Development

The loss of E-cadherin expression is a common feature of EMT (Kalluri and Weinberg,

2009; Thiery et al., 2009). As shown in Figures 3 and 4, Scrib deficiency resulted in reduced

levels of E-cadherin within the epithelial cells of the ScribLecre lenses (Figures 3E-H and 4A).

Because conditional deletion of Cdh1 in the lens placode resulted in an E-cadherin deficient lens

� 15

vesicle and ultimately resulted in EMT (Pontoriero et al., 2009), we asked if loss of Scrib at this

same stage in lens development resulted in changes in E-cadherin localization and/or levels at the

time of lens vesicle formation. Double immunofluorescence experiments were carried out for E-

cadherin and �-catenin on lens sections from embryonic day 11.5 (E11.5) control and

ScribLeCre embryos. At E11.5, control lens vesicles showed E-cadherin (green) colocalized with

�-catenin (red) along lateral membranes with a concentration of E-cadherin at the apical, anterior

surfaces of the lens vesicle (yellow, Figure 5A, B, arrows). In contrast, colocalization was

reduced in this area of ScribLeCre lens vesicles (Figure 5C). An examination of E-cadherin

staining alone suggested that apparent loss of colocalization was due to reduced E-cadherin in

the apical, anterior surface of the Scrib-deficient lens vesicle (Figure 5D). To follow the

temporal progression of E-cadherin disruption, sections from E13.5-E17.5 embryos were

immunostained. By E13.5 and E17.5, colocalization in the controls was restricted to the lateral

and basal membranes of the epithelium (Figure 5E, I). Punctate E-cadherin staining was

observed at the apical surfaces of cells (Figure 5F, J). In contrast, in regions of the E13.5 and

E17.5 ScribLeCre epithelium, staining was red rather than yellow, suggesting that these regions

lacked E-cadherin (Figure 5G, K, arrows). Unmerged images showed that in regions along the

basal surface E-cadherin staining was absent and along the lateral membranes E-cadherin

staining was reduced (Figure 5H, arrows). This trend continued through ScribLeCre E17.5

lenses (Figure 5L, arrows). Alterations in E-cadherin staining of E13.5 and E17.5 ScribLeCre

lenses correlated well with morphological defects in the epithelium. In comparison to controls,

cells in the epithelium of the E13.5 ScribLeCre mice were disorganized (Figure 5G); by E17.5,

the epithelium was obviously flattened and characterized by irregularly shaped cells (Figure 5K).

� 16

Thus, beginning at very early stages of lens formation, the Scrib-deficient lens epithelium is

characterized by the gradual loss of E-cadherin from the epithelium, a hallmark of EMT.

The Apical Marker, ZO-1, is Gradually Lost From Epithelial Cells of the ScribLeCre

Lenses

Another hallmark of EMT is the loss of the apically-restricted tight junction protein, ZO-

1 (Zeisberg and Neilson, 2009). In Cdh1 deficient lenses, both the localization and levels of ZO-

1 were disrupted (Pontoriero et al., 2009). Therefore, we asked if ZO-1 was similarly altered in

the epithelium of ScribLeCre lenses. Paraffin sections from E11.5, E13.5, E15.5, and E17.5

lenses were subjected to immunofluorescence analysis using an anti-ZO-1 antibody. At E11.5

the pattern of ZO-1 staining in ScribLeCre lenses did not differ from that in the controls (data not

shown). However, by E13.5 we observed differences in the pattern of ZO-1 staining. At this

stage, in the epithelium of control lenses ZO-1 localized specifically to the apical membrane of

the epithelial cells, which was evident where the fiber cells had not fully elongated (Figure 6A,

arrows). At E15.5 and E17.5, intense ZO-1 staining was observed at the apical-apical interface

of epithelial and fiber cells (Figure 6D, F arrows). In contrast, by E13.5, in the ScribLeCre

lenses, ZO-1 staining was greatly reduced at the apical surface of the epithelial cells, which was

evident in regions where the fiber cells had not fully elongated (Figure 6B, arrow). By E15.5,

there were regions of ScribLeCre lens epithelium lacking ZO-1 (Figure 6D, arrows). By E17.5

there were regions lacking ZO-1 (Figure 6F, arrows) and, overall, the regular punctuate staining

at the apical-apical interface of the epithelial and fiber cells was largely absent (Figure 6F).

Thus, in the absence of Scrib, ZO-1 is gradually lost from the epithelium, a second hallmark of

EMT.

� 17

Downregulation of Pax6 Correlates Temporally with Upregulation of �SMA

Since we observed defects in E-cadherin and ZO-1 in the embryonic ScribLeCre lenses

beginning at early stages in lens differentiation, we asked at what age loss of Pax6 first would be

observed. Sections from day E13.5, E15.5 and E17.5 control and ScribLeCre lenses were

immunostained for Pax6. At E13.5 and E15.5, no difference was observed in the intensity of

staining between controls and ScribLeCre lenses (data not shown). However, by E17.5, reduced

intensity of staining for Pax6 was observed in some nuclei in the epithelium from ScribLeCre

lenses as compared to controls (Figure 7A-D). We then asked if Pax6 downregulation correlated

temporally with upregulation of �SMA and indicators of TGF� signaling. Sections from E13.5,

E15.5, and E17.5 control and ScribLeCre lenses were immunostained with antibodies against

�SMA and Smad4. At E13.5 or E15.5, no staining for �SMA was observed in the lens

epithelium of control or ScribLeCre mice and Smad4 staining was cytoplasmic (data not shown).

However, at E17.5, staining for �SMA (Figure 7G, H) was observed in ScribLeCre lenses, but

not in controls (Figure 7E, F and I, J). Additionally, and staining for Smad4 was concentrated in

the nuclei in the epithelium of ScribLeCre lenses (Figure 7K, L) TGF� signaling through Smads

requires activation of Smad2/3 by phosphorylation, which forms a nuclear complex with

activated Smad4 (Itoh et al., 2000). Therefore, to further define the signaling pathway that was

now activated in the ScribLeCre lens epithelial cells, we asked if nuclear Smad3 could also be

detected in ScribLeCre lenses. To do so, immunohistochemistry using anti-Smad3 antibodies

was performed on eye sections from E17.5 and P10 control and ScribLeCre animals. No Smad3

staining was observed in control lenses (Figure 8A). However, Smad3 positive nuclei were

observed in ScribLeCre lenses (Figure 8B, arrows). Total and Smad3 positive nuclei were

� 18

counted on 6-8 sections from 3 different eyes. In the ScribLeCre lens epithelia, 21.44% (p=0.02)

of nuclei were positive for Smad3. Similarly, at P10, Smad3 staining was detected in the nuclei

of lens epithelial cells of ScribLeCre eyes but not in the lens epithelial cells of controls (Figure

8C, D, arrows).

TGF� activity induces the expression of zinc finger transcription factors, Snail, Slug, and

Twist, which are transcriptional repressors of Cdh1 (Moustakas and Heldin, 2007). Because

there was obvious loss of E-cadherin from lens epithelia of ScribLeCre mice from E17.5 through

P10 (Figures 3 and 5), we asked if loss of E-cadherin at these time points coincided with the

appearance of Snail in the lens epithelium. To do so, immunohistochemistry using anti-Snail

antibody was performed on eye sections from E17.5 and P10 mice. In E17.5 controls, some

background staining was observed only in the cytoplasm of epithelial cells (Figure 8E).

However, in ScribLeCre lenses, cytoplasmic and nuclear staining was observed (Figure 8F,

arrows). Total and Snail positive nuclei were counted on 6-8 sections from 3 different eyes. In

ScribLeCre lens epithelia 20.87% (p=0.01) of nuclei were positive for Snail. At P10, nuclear

staining for Snail was observed in ScribLeCre lenses but not in control lenses (Figure 8G, H,

arrows). Thus, the appearance of �SMA and TGF� signaling molecules correlate temporally

with the loss of lens epithelial protein, Pax6, and the continued loss of E-cadherin.

Molecular Changes in the Corneal Epithelium of ScribLeCre Mice

As shown in Figure 2E and F, the corneal epithelium in the ScribLeCre mice appeared

disorganized and basal layer cells appeared flattened rather than cuboidal. We have previously

determined that Scrib is expressed in the corneal epithelium (Nguyen et al., 2005) and that Scrib

� 19

protein was not detected in the corneas of P10 ScribLeCre mice (Figure 1E). Given the effect of

loss of Scrib in the lens, we asked if loss of Scrib in the corneal epithelium also resulted in EMT

in that tissue. At P10, the stratified corneal epithelium is characterized by E-cadherin (Figure

9A,B; green), which normally localizes at cell membranes of the basal and superficial layers of

the corneal epithelium (Takahashi et al., 1992). In the ScribLeCre corneal epithelium, E-

cadherin staining was reduced and absent from many cells in the superficial layers (Figure

9C,D). The corneal epithelium is also marked by the uniform expression of cytokeratin K12 in

the suprabasal and superficial layers of the epithelium (Kao et al., 1996). Staining for K12 was

observed throughout all membranes in the epithelium of control corneas at P10 (Figure 9E, F;

red). However, K12 staining in the ScribLeCre corneal epithelium was not uniform, with

regions showing reduced staining (Figure 9G, H), suggesting that the integrity of this epithelium

was compromised. To determine if cells in the corneal epithelium were in transition to a

mesenchymal state, sections were immunostained with anti-�SMA antibodies. Whereas the

corneal epithelium from control mice was negative for �SMA, �SMA staining was present in the

ScribLeCre corneal epithelium and endothelium (Figure 9K, L). Therefore, loss of Scrib is in a

second epithelium, the corneal epithelium, leads to molecular changes consistent with EMT.

DISCUSSION

Epithelial cells are distinguished from other cell types in part by the presence of strong

cell-cell adhesion and apical-basal polarity. The loss of the epithelial proteins, E-cadherin and

ZO-1, major components of adherens and tight junctions, respectively, from epithelial cells and

appearance of mesenchymal proteins such as �SMA and activated TGF� signaling intermediates

are characteristics found in many examples of EMT (Kalluri and Weinberg, 2009; Thiery et al.,

2009). Studies from invertebrate systems indicate that scrib is one factor playing an important

� 20

role in the genesis and maintenance of epithelial adherens junctions and apical polarity

complexes. In this study, using the mouse lens as a model epithelial tissue, we found that loss of

Scrib at the lens placode stage resulted in loss of E-cadherin and ZO-1, components of adherens

junctions and apical polarity complexes, indicating that Scrib plays a conserved role cross

species. Furthermore, the loss of these factors, which are also lost in EMT, was accompanied by

other molecular changes characteristic of EMT (summarized in Figure S1). Molecular changes

consistent with EMT also occurred in a second ocular derivative of the surface ectoderm, the

corneal epithelium. Thus, these data suggest that Scrib may generally be required for establishing

or maintaining the specialized characteristics that define epithelial cells, which are necessary for

maintaining the integrity of an epithelium in vivo and for preventing EMT.

Scrib is an Upstream Regulator of Epithelial Identity in the Lens

In the lens, conditional deletion of Cdhl1 at the lens vesicle stage leads to the loss of cell

adhesion and shape, the gradual loss of ZO-1, and, ultimately, EMT (Pontoriero et al., 2009),

indicating the importance of E-cadherin in maintaining epithelial identity. In ScribLeCre lenses,

the first phenotypic changes we observed were altered cell morphology and a reduced

accumulation of E-cadherin at the apical borders of the cells in lens vesicle. These initial changes

were followed by gradual loss of ZO-1. Given the similarities in the lens phenotype between the

Cdh1 conditional mutant and ScribLeCre mice, we suggest that Scrib functions upstream of E-

cadherin as a regulator of epithelial identity as it is required to maintain normal cell-cell adhesion

and apical polarity, two defining characteristics of epithelial cells.

� 21

Interestingly, the developmental time point at which we first observed reduced

accumulation of E-cadherin, E11.5, is only one day after the lens vesicle normally separates from

the overlying ectoderm. For detachment of the lens vesicle from the overlying ectoderm to occur,

E-cadherin-containing cell-cell adherens junctions must be broken and subsequently reformed

between new neighboring cells (Wiley et al., 2010). It is possible, therefore, that reestablishing

adherens junctions to form the lens vesicle is impaired in the absence of Scrib. In keeping with

this concept, deletion of Scrib after the lens vesicle has formed does not result in EMT

(unpublished data), suggesting Scrib may be required for unique events involving E-cadherin

function at the time of lens vesicle formation. The loss of ZO-1 from the apical domain of the

epithelial cells beginning at E13.5 suggests that disrupting adherens junctions leads to the

destabilization of the tight junction complexes. Similarly, reduced levels of another apical

polarity protein, Par3, which is part of the Par3/Par6/aPKC complex (Humbert et al., 2008;

Wodarz and Nathke, 2007), was observed on the central epithelial cells of ScribLeCre lens and

in the transition zone Par3 was also was observed on the lateral membranes (data not shown),

indicating that restriction of Par3 to the apical domain was lost. This is reminiscent of findings in

Drosophila where loss of lgl, a component of the scrib polarity complex, leads to expansion of

the apical domain into the basolateral domain (Kaplan et al., 2009). Our data suggest that the

dynamic stability of these complexes in vivo in the mouse lens epithelium requires Scrib and that

these are required to maintain epithelial identity. However, Scrib may also be responsible at the

time of lens vesicle formation and shortly thereafter for other cellular and molecular events that

are crucial for forming and maintaining a lens epithelium. For example, we have noticed that

ScribLeCre lenses show defects cell viability and proliferation (Yamben and Griep, manuscript

in preparation). These changes could, along with the fiber cell defects, contribute to the overall

� 22

phenotype of the ScribLeCre lens. Additional investigations are needed to further elucidate

Scrib’s involvement with early lens formation and how this impacts on maintaining epithelial

identity as well as the overall integrity of the lens.

Following initial disruption in E-cadherin localization and loss of ZO-1, we observed

evidence for active TGF� signaling and the appearance of, �SMA. Thus, further decreases in E-

cadherin expression observed in late embryonic and postnatal ScribLeCre lenses may be the

result of Snail activity, as Snail is a known negative regulator of Cdh1 transcription (Moustakas

and Heldin, 2007). Loss of the lens epithelial transcription factor, Pax6, also was observed. This

likely contributes to the loss of lens epithelial identity in the Scrib-deficient lenses, since lenses

of mice haplo-insufficient for Pax6 exhibit EMT (Lovicu et al., 2004) and Pax6 is a regulator of

expression of the lens specific crystallin genes (Cvekl et al., 2004). It has been shown that

downregulation of Pax6 occurs in the TGF� overexpressing transgenic mouse lens (Lovicu et al.,

2004) and activated Smad3 can negatively regulate Pax6 expression (Grocott et al., 2007). The

timepoint (E17.5) when Pax6 loss was first noticed in the ScribLeCre lenses suggests TGF�

signaling may be responsible. Thus, we suggest that in the Scrib deficient lenses, EMT initiates

with the destabilization of the junctional complexes and apical-basal polarity and then is driven

through the acquisition of active TGF��signaling.

Scrib, EMT and Disease

The loss of epithelial identity and EMT is associated with two types of cataract in human

and mouse models, ASC (Hales et al., 1995; Lovicu et al., 2004; Saika et al., 2009; Srinivasan et

al., 1998) and PCO, (Wormstone et al., 2006; Wormstone et al., 2002; Wormstone et al., 2009),

which commonly occurs after cataract surgery. Phenotypically, the lenses of ScribLeCre mice

� 23

mimic many of the defects observed in these models of cataract. Additionally, defects in fiber

cell adhesion have been shown to be a contributing factor to cortical cataract in experimental

models (Zhou et al., 2007). ScribLeCre lenses also exhibit defects in cell adhesion in the fiber

cell compartment suggesting that Scrib function may be required to maintain normal fiber cell

characteristics and prevent cortical cataractogenesis. It is not known if Scrib expression and/or

function is compromised in any of these examples of mouse or human cataract. Further studies

will be required to address this question.

The loss of Scrib led to EMT not only in the lens but also in the corneal epithelium, as

evidenced by the reduced corneal epithelial proteins, E-cadherin and K12, and the expression of

�SMA. This suggests that Scrib may be required to maintain epithelial identity, thus preventing

EMT and disease in many different organs. In the eye, EMT of corneal epithelial cells in the

limbal region has been suggested to be involved in the pathogenesis of pterygium (Kato et al.,

2007). EMT of retinal pigment epithelial cells is observed in proliferative vitreoretinopathy, a

consequence of retinal detachment (Hiscott et al., 1999; Nagasaki et al., 1998; Thiery et al.,

2009). EMT is associated with tissue fibrosis in a variety of other organs such as kidney, liver,

and lung (Kalluri and Weinberg, 2009). EMT and the targeting of cell polarity complexes by

EMT inducers have been associated with increased invasiness and metastasis in cancer (Banks et

al., 2012; Elsum et al., 2012; Moreno-Bueno et al., 2008; Olmeda et al., 2008). For example, the

human papillomavirus oncogenes have been shown to downregulate epithelial proteins while

upregulating mesenchymal proteins in human keratinocytes (Hellner et al., 2009). Of particular

relevance to our study, expression of the E6 oncoprotein, dependent on an intact PDZ binding

domain through which E6 binds to and promotes the degradation of Scrib (Thomas et al., 2005),

has been shown to result in phenotypic changes in cultured keratinocytes that are indicative of

� 24

EMT (Watson et al., 2003). In cervical cancer specimens, greatly reduced levels of Scrib are

observed in high-grade squamous intraepithelial lesions (SIL) and invasive cancers as compared

to low grade SIL lesions (Nakagawa et al., 2004), showing an inverse correlation between Scrib

levels and the degree of invasiveness of the tumor. Also, reduced levels of Scrib are associated

with loss of epithelial polarity and tissue architecture in colon adenocarcinoma (Gardiol et al.,

2006), metastasis in endometrial cancer (Ouyang et al., 2010), and breast cancer (Zhan et al.,

2008). Thus, it is possible that the loss of Scrib function results in EMT that drives the

metastatic progression of many types of cancers.

CONCLUSIONS

Together, these data suggest that, in vivo, mammalian Scrib is required to maintain

epithelial identity and its loss allows epithelial cells to be susceptible to EMT. Scrib may

maintain epithelial identity primarily through its effect on junctional complexes. An examination

of molecular changes during the early stages of lens formation in ScribLeCre lenses may provide

further insight into roles of Scrib in EMT and human disease.

ACKNOWLEDGEMENTS

The authors thank Ruth Ashery-Padan for generously providing the LeCre mice. The authors

also thank Lance Roderick of the Keck Imaging Facility, Toshi Kinoshita of the Pathology

Department Histology Core, the McArdle Laboratories Histology Core and Denis Lee and Susan

Moran of the McArdle Laboratories. The authors thank Angela Verdoni and Aki Ikeda for

guidance with the cornea experiments. The authors thank Paul Lambert for helpful discussions

and his critical reading of the manuscript. I.F.Y. was supported by the NIH training grant

� 25

GM07215. This work was supported by NIH grants EY09091, CA98428, CA14520, and

EY016665.

REFERENCES

Acloque, H., Adams, M.S., Fishwick, K., Bronner-Fraser, M., Nieto, M.A., 2009. Epithelial-mesenchymal transitions: the importance of changing cell state in development and disease. J Clin Invest 119, 1438-1449. Ashery-Padan, R., Marquardt, T., Zhou, X., Gruss, P., 2000. Pax6 activity in the lens primordium is required for lens formation and for correct placement of a single retina in the eye. Genes Dev 14, 2701-2711. Banks, L., Pim, D., Thomas, M., 2012. Human tumour viruses and the deregulation of cell polarity in cancer. Nat Rev Cancer 12, 877-886. Baum, B., Georgiou, M., 2011. Dynamics of adherens junctions in epithelial establishment, maintenance, and remodeling. J Cell Biol 192, 907-917. Bilder, D., 2004. Epithelial polarity and proliferation control: links from the Drosophila neoplastic tumor suppressors. Genes Dev 18, 1909-1925. Bilder, D., Li, M., Perrimon, N., 2000. Cooperative regulation of cell polarity and growth by Drosophila tumor suppressors. Science 289, 113-116. Bilder, D., Perrimon, N., 2000. Localization of apical epithelial determinants by the basolateral PDZ protein Scribble. Nature 403, 676-680. Cvekl, A., Yang, Y., Chauhan, B.K., Cveklova, K., 2004. Regulation of gene expression by Pax6 in ocular cells: a case of tissue-preferred expression of crystallins in lens. Int J Dev Biol 48, 829-844. de Iongh, R.U., Wederell, E., Lovicu, F.J., McAvoy, J.W., 2005. Transforming growth factor-beta-induced epithelial-mesenchymal transition in the lens: a model for cataract formation. Cells Tissues Organs 179, 43-55. Dwivedi, D.J., Pino, G., Banh, A., Nathu, Z., Howchin, D., Margetts, P., Sivak, J.G., West-Mays, J.A., 2006. Matrix metalloproteinase inhibitors suppress transforming growth factor-beta-induced subcapsular cataract formation. Am J Pathol 168, 69-79. Elsum, I., Yates, L., Humbert, P.O., Richardson, H.E., 2012. The Scribble-Dlg-Lgl polarity module in development and cancer: from flies to man. Essays Biochem 53, 141-168.

� 26

Gardiol, D., Zacchi, A., Petrera, F., Stanta, G., Banks, L., 2006. Human discs large and scrib are localized at the same regions in colon mucosa and changes in their expression patterns are correlated with loss of tissue architecture during malignant progression. Int J Cancer 119, 1285-1290. Giepmans, B.N., van Ijzendoorn, S.C., 2009. Epithelial cell-cell junctions and plasma membrane domains. Biochim Biophys Acta 1788, 820-831. Grocott, T., Frost, V., Maillard, M., Johansen, T., Wheeler, G.N., Dawes, L.J., Wormstone, I.M., Chantry, A., 2007. The MH1 domain of Smad3 interacts with Pax6 and represses autoregulation of the Pax6 P1 promoter. Nucleic Acids Res 35, 890-901. Hales, A.M., Chamberlain, C.G., McAvoy, J.W., 1995. Cataract induction in lenses cultured with transforming growth factor-beta. Invest Ophthalmol Vis Sci 36, 1709-1713. Harris, B.Z., Lim, W.A., 2001. Mechanism and role of PDZ domains in signaling complex assembly. J Cell Sci 114, 3219-3231. Hellner, K., Mar, J., Fang, F., Quackenbush, J., Munger, K., 2009. HPV16 E7 oncogene expression in normal human epithelial cells causes molecular changes indicative of an epithelial to mesenchymal transition. Virology 391, 57-63. Hiscott, P., Sheridan, C., Magee, R.M., Grierson, I., 1999. Matrix and the retinal pigment epithelium in proliferative retinal disease. Prog Retin Eye Res 18, 167-190. Humbert, P., Russell, S., Richardson, H., 2003. Dlg, Scribble and Lgl in cell polarity, cell proliferation and cancer. Bioessays 25, 542-553. Humbert, P.O., Grzeschik, N.A., Brumby, A.M., Galea, R., Elsum, I., Richardson, H.E., 2008. Control of tumourigenesis by the Scribble/Dlg/Lgl polarity module. Oncogene 27, 6888-6907. Itoh, S., Itoh, F., Goumans, M.J., Ten Dijke, P., 2000. Signaling of transforming growth factor-beta family members through Smad proteins. Eur J Biochem 267, 6954-6967. Johnson, K., Wodarz, A., 2003. A genetic hierarchy controlling cell polarity. Nat Cell Biol 5, 12-14. Kalluri, R., Weinberg, R.A., 2009. The basics of epithelial-mesenchymal transition. J Clin Invest 119, 1420-1428. Kao, W.W., Liu, C.Y., Converse, R.L., Shiraishi, A., Kao, C.W., Ishizaki, M., Doetschman, T., Duffy, J., 1996. Keratin 12-deficient mice have fragile corneal epithelia. Invest Ophthalmol Vis Sci 37, 2572-2584.

� 27

Kaplan, N.A., Liu, X., Tolwinski, N.S., 2009. Epithelial polarity: interactions between junctions and apical-basal machinery. Genetics 183, 897-904. Kato, N., Shimmura, S., Kawakita, T., Miyashita, H., Ogawa, Y., Yoshida, S., Higa, K., Okano, H., Tsubota, K., 2007. Beta-catenin activation and epithelial-mesenchymal transition in the pathogenesis of pterygium. Invest Ophthalmol Vis Sci 48, 1511-1517. Klymkowsky, M.W., Savagner, P., 2009. Epithelial-mesenchymal transition: a cancer researcher's conceptual friend and foe. Am J Pathol 174, 1588-1593. Lewandoski, M., Meyers, E.N., Martin, G.R., 1997. Analysis of Fgf8 gene function in vertebrate development. Cold Spring Harb Symp Quant Biol 62, 159-168. Lovicu, F.J., Ang, S., Chorazyczewska, M., McAvoy, J.W., 2004. Deregulation of lens epithelial cell proliferation and differentiation during the development of TGFbeta-induced anterior subcapsular cataract. Dev Neurosci 26, 446-455. Michaelis, U.R., Chavakis, E., Kruse, C., Jungblut, B., Kaluza, D., Wandzioch, K., Manavski, Y., Heide, H., Santoni, M.J., Potente, M., Eble, J.A., Borg, J.P., Brandes, R.P., 2013. The polarity protein scrib is essential for directed endothelial cell migration. Circ Res 112, 924-934. Montcouquiol, M., Rachel, R.A., Lanford, P.J., Copeland, N.G., Jenkins, N.A., Kelley, M.W., 2003. Identification of Vangl2 and Scrb1 as planar polarity genes in mammals. Nature 423, 173-177. Moreno-Bueno, G., Portillo, F., Cano, A., 2008. Transcriptional regulation of cell polarity in EMT and cancer. Oncogene 27, 6958-6969. Moustakas, A., Heldin, C.H., 2007. Signaling networks guiding epithelial-mesenchymal transitions during embryogenesis and cancer progression. Cancer Sci 98, 1512-1520. Murdoch, J.N., Henderson, D.J., Doudney, K., Gaston-Massuet, C., Phillips, H.M., Paternotte, C., Arkell, R., Stanier, P., Copp, A.J., 2003. Disruption of scribble (Scrb1) causes severe neural tube defects in the circletail mouse. Hum Mol Genet 12, 87-98. Nagasaki, H., Shinagawa, K., Mochizuki, M., 1998. Risk factors for proliferative vitreoretinopathy. Prog Retin Eye Res 17, 77-98. Nakagawa, S., Yano, T., Nakagawa, K., Takizawa, S., Suzuki, Y., Yasugi, T., Huibregtse, J.M., Taketani, Y., 2004. Analysis of the expression and localisation of a LAP protein, human scribble, in the normal and neoplastic epithelium of uterine cervix. Br J Cancer 90, 194-199. Nakaya, Y., Sheng, G., 2008. Epithelial to mesenchymal transition during gastrulation: an embryological view. Dev Growth Differ 50, 755-766.

� 28

Nguyen, M.M., Rivera, C., Griep, A.E., 2005. Localization of PDZ domain containing proteins Discs Large-1 and Scribble in the mouse eye. Mol Vis 11, 1183-1199. Olmeda, D., Montes, A., Moreno-Bueno, G., Flores, J.M., Portillo, F., Cano, A., 2008. Snai1 and Snai2 collaborate on tumor growth and metastasis properties of mouse skin carcinoma cell lines. Oncogene 27, 4690-4701. Ouyang, Z., Zhan, W., Dan, L., 2010. hScrib, a human homolog of Drosophila neoplastic tumor suppressor, is involved in the progress of endometrial cancer. Oncol Res 18, 593-599. Pearson, H.B., Perez-Mancera, P.A., Dow, L.E., Ryan, A., Tennstedt, P., Bogani, D., Elsum, I., Greenfield, A., Tuveson, D.A., Simon, R., Humbert, P.O., 2011. SCRIB expression is deregulated in human prostate cancer, and its deficiency in mice promotes prostate neoplasia. J Clin Invest 121, 4257-4267. Piatigorsky, J., 1981. Lens differentiation in vertebrates. A review of cellular and molecular features. Differentiation 19, 134-153. Pontoriero, G.F., Smith, A.N., Miller, L.A., Radice, G.L., West-Mays, J.A., Lang, R.A., 2009. Co-operative roles for E-cadherin and N-cadherin during lens vesicle separation and lens epithelial cell survival. Dev Biol 326, 403-417. Qin, Y., Capaldo, C., Gumbiner, B.M., Macara, I.G., 2005. The mammalian Scribble polarity protein regulates epithelial cell adhesion and migration through E-cadherin. J Cell Biol 171, 1061-1071. Rivera, C., Yamben, I.F., Shatadal, S., Waldof, M., Robinson, M.L., Griep, A.E., 2009. Cell-autonomous requirements for Dlg-1 for lens epithelial cell structure and fiber cell morphogenesis. Dev Dyn 238, 2292-2308. Saika, S., Yamanaka, O., Flanders, K.C., Okada, Y., Miyamoto, T., Sumioka, T., Shirai, K., Kitano, A., Miyazaki, K., Tanaka, S., Ikeda, K., 2008. Epithelial-mesenchymal transition as a therapeutic target for prevention of ocular tissue fibrosis. Endocr Metab Immune Disord Drug Targets 8, 69-76. Saika, S., Yamanaka, O., Okada, Y., Tanaka, S., Miyamoto, T., Sumioka, T., Kitano, A., Shirai, K., Ikeda, K., 2009. TGF beta in fibroproliferative diseases in the eye. Front Biosci (Schol Ed) 1, 376-390. Srinivasan, Y., Lovicu, F.J., Overbeek, P.A., 1998. Lens-specific expression of transforming growth factor beta1 in transgenic mice causes anterior subcapsular cataracts. J Clin Invest 101, 625-634. Sugiyama, Y., Akimoto, K., Robinson, M.L., Ohno, S., Quinlan, R.A., 2009. A cell polarity protein aPKClambda is required for eye lens formation and growth. Dev Biol 336, 246-256.

� 29

Symonds, J.G., Lovicu, F.J., Chamberlain, C.G., 2006. Posterior capsule opacification-like changes in rat lens explants cultured with TGFbeta and FGF: effects of cell coverage and regional differences. Exp Eye Res 82, 693-699. Takahashi, M., Fujimoto, T., Honda, Y., Ogawa, K., 1992. Distributional change of fodrin in the wound healing process of the corneal epithelium. Invest Ophthalmol Vis Sci 33, 280-285. Thiery, J.P., Acloque, H., Huang, R.Y., Nieto, M.A., 2009. Epithelial-mesenchymal transitions in development and disease. Cell 139, 871-890. Thomas, M., Massimi, P., Navarro, C., Borg, J.P., Banks, L., 2005. The hScrib/Dlg apico-basal control complex is differentially targeted by HPV-16 and HPV-18 E6 proteins. Oncogene 24, 6222-6230. Tsuji, T., Ibaragi, S., Hu, G.F., 2009. Epithelial-mesenchymal transition and cell cooperativity in metastasis. Cancer Res 69, 7135-7139. Watson, R.A., Thomas, M., Banks, L., Roberts, S., 2003. Activity of the human papillomavirus E6 PDZ-binding motif correlates with an enhanced morphological transformation of immortalized human keratinocytes. J Cell Sci 116, 4925-4934. Wiley, L.A., Dattilo, L.K., Kang, K.B., Giovannini, M., Beebe, D.C., 2010. The tumor suppressor merlin is required for cell cycle exit, terminal differentiation, and cell polarity in the developing murine lens. Invest Ophthalmol Vis Sci 51, 3611-3618. Wodarz, A., Nathke, I., 2007. Cell polarity in development and cancer. Nat Cell Biol 9, 1016-1024. Woods, D.F., Hough, C., Peel, D., Callaini, G., Bryant, P.J., 1996. Dlg protein is required for junction structure, cell polarity, and proliferation control in Drosophila epithelia. J Cell Biol 134, 1469-1482. Wormstone, I.M., Anderson, I.K., Eldred, J.A., Dawes, L.J., Duncan, G., 2006. Short-term exposure to transforming growth factor beta induces long-term fibrotic responses. Exp Eye Res 83, 1238-1245. Wormstone, I.M., Tamiya, S., Anderson, I., Duncan, G., 2002. TGF-beta2-induced matrix modification and cell transdifferentiation in the human lens capsular bag. Invest Ophthalmol Vis Sci 43, 2301-2308. Wormstone, I.M., Wang, L., Liu, C.S., 2009. Posterior capsule opacification. Exp Eye Res 88, 257-269. Yates, L.L., Schnatwinkel, C., Hazelwood, L., Chessum, L., Paudyal, A., Hilton, H., Romero, M.R., Wilde, J., Bogani, D., Sanderson, J., Formstone, C., Murdoch, J.N., Niswander, L.A.,

� 30

Greenfield, A., Dean, C.H., 2013. Scribble is required for normal epithelial cell-cell contacts and lumen morphogenesis in the mammalian lung. Dev Biol 373, 267-280. Zeisberg, M., Neilson, E.G., 2009. Biomarkers for epithelial-mesenchymal transitions. The Journal of clinical investigation 119, 1429-1437. Zhan, L., Rosenberg, A., Bergami, K.C., Yu, M., Xuan, Z., Jaffe, A.B., Allred, C., Muthuswamy, S.K., 2008. Deregulation of scribble promotes mammary tumorigenesis and reveals a role for cell polarity in carcinoma. Cell 135, 865-878. Zhao, M., Szafranski, P., Hall, C.A., Goode, S., 2008. Basolateral junctions utilize warts signaling to control epithelial-mesenchymal transition and proliferation crucial for migration and invasion of Drosophila ovarian epithelial cells. Genetics 178, 1947-1971. Zhou, J., Leonard, M., Van Bockstaele, E., Menko, A.S., 2007. Mechanism of Src kinase induction of cortical cataract following exposure to stress: destabilization of cell-cell junctions. Mol Vis 13, 1298-1310.

FIGURE LEGENDS

Figure 1. Generation of Scrib conditional knockout mice. A: Shown are exons 1-10 of the

wildtype (WT) Scrib allele. The targeting allele (TA) shown illustrates loxP sites that flank

exons 2-8 and frt sites that flank a positive selection neomycin (NEO) marker. Mice containing

the targeting allele were mated to mice expressing Flpase, which removed the neo marker and

generated the floxed (F) allele. Arrows indicate locations of PCR primers a (ScribF), b

(ScribwtR), and c (ScribckoR) used to identify if exons 2-8 had been deleted and d (Scrib1-X2-

R) used in combination with a to routinely genotype the floxed and wt alleles. The mutant allele

was generated through mating with Lens-Cre. B: Immunofluoresence demonstrating localization

of Scrib (red) in the E10.5 lens vesicle (lv). Arrow indicates concentrated apical localization in

the lens vesicle. oc, optic cup. C: GFP fluorescence indicating cre activity specifically in the

E10.5 ScribLecre vesicle. D: Lens specific cre-mediated deletion of exons 2-8. PCR primer a,

� 31

b, and c were used to amplify DNA fragments from lens, eye and tail samples. The wt

band=437bp, the null band=193bp and the floxed band=325bp. ScribLeCre lenses had the 193bp

null band. E: Whole cell lysates from lenses of P2 and corneas of P10 control and ScribLeCre

mice were subjected to immunoblot analysis for Scrib and reprobed for Gapdh as a loading

control. Scrib was reduced in lenses from ScribLeCre mice as compared controls and absent in

the cornea lysate from ScribLeCre mice. Protein lysates from Scrib wt and null brain were

included to demonstrate the specificity of the antibody.

Figure 2: Morphological defects in the eyes of postnatal day 10 (P10) ScribLeCre mice. A-

B: Eyes (A) and lenses (B) were isolated from control and ScribLeCre mice and viewed under a

dissecting microscope. A: P10 ScribLeCre eyes were noticeably smaller (right) compared to

controls (left). B: ScribLeCre lenses were smaller than controls and also had an opaque center

indicating a cataract (arrow). C-J: Longitudinally oriented paraffin embedded sections of eyes

from P10 control (C, E, G) and ScribLeCre (D, F, H) mice were stained with hematoxylin and

eosin. C-D: Sections of controls (C) and ScribLeCre eyes (D) showing that mutant lenses were

vacuolated (arrow) and nuclei were scattered throughout the lens (arrowhead). Also, the iris was

hyperplastic (asterisk). E-F: Higher magnification of the lens epithelium in control (E) and

ScribLeCre (F) lenses highlighting the flattened and elongated cells in the epithelium of the

mutant as compared to controls (arrowheads). G-H: Higher magnification image of control (G)

and ScribLeCre (H) corneas showing that the epithelium of mutant corneas lacked an organized

stratified epithelium with cuboidal shaped cells in the basal layer (arrowheads). Insets show

higher magnification images of the regions indicated by black boxes. ac, anterior chamber; c,

� 32

cornea; e, epithelium; en, endothelium; f, fiber cells; i, iris; l, lens; s, stroma. Bar = 50μm for

C, D and 25μm for E-H.

Figure 3: Epithelial proteins are downregulated and mesenchymal proteins are

upregulated in the P10 ScribLeCre lens epithelium. Longitudinally oriented, paraffin

embedded (A-H, M-P) or cryogenic (I-L) sections from P10 control and ScribLeCre eyes were

immunostained with antibodies against epithelial proteins (Pax6; A-D and E-Cadherin; E-H), the

mesenchymal protein, �SMA (I-L), and Smad4 (M-P). Sections were counterstained with either

propidium iodide (red, A, C), To-Pro3 (blue, I, K, M, O) or anti-�-catenin antibodies (E, G). A-

D: Pax6 (green) was found consistently in the nuclei of epithelial cells from control lenses (A-

B) whereas nuclear Pax6 staining was absent from (arrows; C, D) or reduced in nuclei of

ScribLeCre epithelial cells. E-F: E-cadherin (green) colocalized (yellow) with �-catenin (red)

along all membranes of control epithelium. G-H: In the flattened ScribLeCre epithelium there

were areas where E-cadherin and �-catenin were absent from basal and apical membranes

(arrows). I-L: �SMA was observed only in the iris (i) of control lenses (I, L), but was found in

both the iris and the epithelium (arrow) of ScribLeCre lenses (K-L). The red box inset (L) shows

a higher magnification image of the region indicated by the white box. M-P: Smad4 (red) was

cytoplasmic in control epithelial cells (N) but was concentrated in the nuclei of ScribLeCre

epithelial cells (O, P, arrows). e, epithelium; f, fiber cells; i, iris. Bar = 50μm.

Figure 4. E-cadherin and Pax6 levels are reduced in ScribLeCre lenses. Protein lysates from

P2 control and ScribLeCre mice were subjected to western blot analysis using anti-E-cadherin

� 33

(A) and anti-Pax6 (B) antibodies. Blots were reprobed for Gapdh as a loading control. For E-

cadherin, three independent pools of protein were analyzed on three different blots. For Pax6 two

independent pools were analyzed on two different blots. Bands were quantified by

phosphorimager analysis and for E-cadherin statistical analysis was conducted using the two-

sided Wilcoxon Rank Sum test. The levels of E-Cadherin were reduced 35% (p=0.04) in

ScribLeCre lenses as compared to controls. The levels of Pax6 were reduced 55% in ScribLeCre

lenses as compared to controls.

Figure 5: E-cadherin is progressively lost from ScribLeCre lenses. Longitudinally oriented,

paraffin embedded sections from E11.5 (A-D), E13.5 (E-H) and E17.5 (I-L) control (A, B, E, F,

I, J) and ScribLeCre (C, D, G, H, K, L) embryos (A-H) and eyes (I-L) were immunostained with

anti-E-cadherin (green) and anti-�-catenin (red) antibodies. A-B: Colocalization (yellow) was

prominent at the apical surface of cells in the anterior lens vesicle (A, arrow). E-cadherin was

punctate in this region (B). C-D: E-cadherin staining was reduced in the anterior region of the

ScribLeCre lens vesicles (arrows). E-F: Colocalization was observed along the basal and lateral

membranes of controls at E13.5 (E). E-cadherin was punctate along the apical surface (F,

arrow). G-H: In E13.5 ScribLeCre lenses, E-cadherin was specifically reduced along the basal

and lateral membranes surface (arrows). I-J: In E17.5 control lenses, colocalization was found

along all surfaces of the epithelium (I). E-cadherin was punctate at the apical surface (J). K-L:

In E17.5 ScribLeCre lenses, E-cadherin was absent or reduced along some basal, lateral

membranes (arrows). Red boxes show higher magnification images of the regions indicated by

white boxes. ce, corneal epithelium; le, lens epithelium; lf, fiber cells; lv, lens vesicle; s, corneal

stroma. Bar= 50 μm.

� 34

Figure 6: The tight junction protein, ZO-1, is gradually lost from the apical surface of

ScribLeCre lens epithelial cells. Longitudinally oriented, paraffin embedded sections of E13.5

embryos (A, B), E15.5 heads (C, D), and E17.5 eyes (E, F) from control (A, C, E) and

ScribLeCre (B, D, F) mice were immunostained with anti-ZO-1 antibodies (green) and the nuclei

counterstained with propidium iodide (red). A-B: At E13.5, ZO-1 staining was observed at the

apical surfaces of epithelial cells in the control lenses (arrows) but was reduced or absent (B,

arrow) on the apical surface of the ScribLeCre lenses. C-D: At E15.5, ZO-1 staining was

observed at the epithelial-fiber interface in control lenses (C, arrows) but was reduced or absent

(D, arrows) on the apical surface of the epithelial cells of the ScribLeCre lenses. E-F: At E17.5,

punctate ZO-1 staining was observed on the apical membrane of the epithelial cells at the

epithelial-fiber interface (E, arrows) but was further reduced or absent (F, arrows) on the apical

membrane of the epithelial cells of the ScribLeCre lenses. The staining pattern in the lens fiber

cells was also disrupted at E17.5. e, epithelium; f, fiber cells. Bar= 50 μm.

Figure 7: Downregulation of Pax6 and upregulation of �SMA and nuclear Smad4 are

observed in the lens epithelium of E17.5 ScribLeCre embryos. Longitudinally oriented,

paraffin embedded (A-D, I-L) or cryogenic (E-H) sections from eyes of E17.5 control (A, B, E,

F, I, J) and ScribLeCre (C, D, G, H, K, L) embryos were immunostained for Pax6 (A-D, green),

�SMA (E-H, red) or Smad4 (I-L, red) and the nuclei counterstained with propidium idodide (A,

C, red) or To-Pro3 (E, G, I, K, blue). A-D: Pax6 staining was observed uniformly in the nuclei of

lens epithelium from control mice (A, B) whereas the intensity of Pax6 staining was variable in

� 35

the nuclei of the epithelium from ScribLeCre mice with some nuclei exhibiting markedly

reduced staining (C, D) and some of these nuclei appeared flattened rather than rounded (C, D,

arrowhead). E-H: Staining for �SMA was observed in the iris but not the lens of control eyes (E,

F) whereas staining was observed in both the lens epithelium (arrows) and iris of the ScribLeCre

eyes (G, H). Note that the iris does not extend across the entire ScribLeCre eye at E17.5 as it

does in P10 ScribLeCre eyes (see Figure 2D). The hyperplastic iris was only observed in eyes of

postnatal ScribLeCre mice. Red boxes show higher magnification images of regions indicated by

white boxes. I-L: Smad4 staining (red) was cytoplasmic in control epithelial cells (I, J) whereas

it was concentrated in the nuclei of some of the ScribLeCre epithelial cells (arrows) and the

nuclei appeared flattened rather than rounded (K, L). e, epithelium f, fibers i, iris. Bar = 50μm.

Figure 8: Nuclear Smad3 and Snail are detected in lens epithelium of ScribLeCre mice.

Longitudinally oriented, paraffin embedded eye sections from control (A, C, E, G) and

ScribLeCre (B, D, F, H) E17.5 embryos (A, B, E, F) or P10 mice (C, D, G, H) were

immunostained for Smad3 (A-D) or Snail (E-H) and counterstained with hematoxylin. A-D:

Nuclear immunoreactivity for Smad3 was not observed in the lens epithelium of E17.5 (A) or

P10 (C) control eyes whereas nuclear Smad3 staining was observed in some nuclei of the lens

epithelium of E17.5 (B, arrows) and P10 (D, arrows) ScribLeCre mice. E-H: Nuclear

immunoreactivity for Snail was not observed in nuclei of E17.5 (E) or P10 (G) control lens

epithelium whereas nuclear staining was observed in ScribLeCre lenses at both E17.5 (F, arrows)

and P10 (H, arrows). ce, corneal epithelium, le, lens epithelium; lf, lens fiber cells; s, corneal

stroma. Bar = 25μm.

� 36

Figure 9: The corneal epithelium in the ScribLeCre mice acquires EMT characteristics.

Longitudinally oriented, cryogenic sections from P10 control and ScribLeCre mice were

immunostained for E-cadherin (A-D, green), cytokeratin 12 (K12, E-H, red) and �SMA (I-L,

red). Nuclei were counterstained with propidium iodide (red, A, C) or To-Pro3 (blue, E, G, I-L).

A-D: E-cadherin (green) is strongly expressed along all cell surfaces of the corneal epithelial

cells of controls (A-B), but is weakly expressed or absent in the corneal epithelial cells from the

ScribLeCre mice (D arrows). E-H: K12 (red) is expressed throughout the corneal epithelium of

control mice but is reduced and discontinuous in corneal epithelium of ScribLeCre mice. (G, H,

arrows). I-L: �SMA (red) was expressed only in ScribLeCre corneal epithelium and

endothelium (K-L, arrows). Red boxes show higher magnification images of epithelium

indicated by the white boxes. e, epithelium; s, stroma; en, endothelium Bars=50μm.

Figure 1

Figu

re 2

Figu

re 3

Figu

re 4

Figu

re 5

Figure 6

Figu

re 7

Figu

re 8

Figu

re 9