SPECIAL ISSUE REVIEWS–A PEER REVIEWED FORUM

Embryonic Chick Corneal Epithelium: A ModelSystem for Exploring Cell–matrix InteractionsKathy K.H. Svoboda,1* Donald A. Fischman,2 and Marion K. Gordon3

In her initial research, Elizabeth D. Hay studied amphibian limb regeneration, but later switched her focus,and for the remainder of her career addressed the role of the extracellular matrix (ECM) in regulatingembryonic morphogenesis. Much of that work used the embryonic chick corneal epithelial model. Thisreview highlights many of the discoveries that she made using this model. Hay was the first to show thatembryonic corneal epithelial cells produce fibrillar collagen. Her lab was among the first to demonstratethat corneal epithelial cells respond to a collagenous substrate by increasing ECM production, and thatpurified ECM molecules, added to cultures of epithelial sheets, induce a reorganization of the actincytoskeleton. These data led to the first theories of cell–matrix interactions, illustrated in a ‘hands acrossthe membrane’ sketch drawn by Hay. Recent work with the epithelial sheet model system has elucidatedmany of the signal transduction pathways required for actin reorganization in response to the ECM. In all,this body of work has amply supported Hay’s belief that the embryonic corneal epithelium is a powerfulmodel system for exploring the role of the ECM in regulating the cytoskeleton, in directing cell migration,and in profoundly influencing cell growth and differentiation during development. Developmental Dynam-ics 237:2667–2675, 2008. © 2008 Wiley-Liss, Inc.

Key words: embryonic corneal epithelium; cell–matrix interactions; cytoskeleton; actin; extracellular matrix; signaltransduction

Accepted 29 May 2008

FIBRILLAR COLLAGENPRODUCTION BYEPITHELIAL CELLS

Elizabeth D. Hay, called Betty, by allwho knew her, was a pioneer in celland developmental biology. She beganher career investigating limb regener-ation in salamanders, and when elec-tron microscopy was being integratedwith autoradiography, she was amongthe first to fruitfully apply them tostudy regeneration. In several earlypapers, Hay reported that both the

salamander epithelial and fibroblastcells of regenerating limbs appearedto be producing, organizing, and as-sembling collagen matrices (Hay andRevel, 1963). At that time, the im-portance of the extracellular matrix(ECM) was just starting to be appre-ciated, and the word “collagen” re-ferred to the only known collagen:fibrillar type I collagen. Hay’s andRevel’s conclusions about the syn-thesis of salamander epithelial typeI collagen were based on morpholog-ical, “pulse chase” experiments. Sys-

temic injection of tritiated prolinedemonstrated that, shortly after in-jection, 3H was localized within in-tracellular organelles, mainly theER and the Golgi apparatus. How-ever, at later time points, the labelwas found in the ECM underlyingthe epithelium, then referred to asthe basement lamella (Hay andRevel, 1963). Publication of thiswork elicited substantial contro-versy, because it went against thecurrent dogma of that time, whichwas that collagen (type I) was pro-

1Department of Biomedical Science, Texas A & M Health Science Center, Baylor College of Dentistry, Dallas, Texas2Department of Cell and Developmental Biology, Weill Medical College of Cornell University, New York, New York3Department of Pharmacology and Toxicology, Ernest Mario School of Pharmacy, Rutgers University, Piscataway, New JerseyGrant sponsor: NIH/NEI; Grant number: EY08886; Grant number: EY014389; Grant number: EY09056; Grant sponsor: UMDNJ/RutgersCenter; Grant number: AR055073.*Correspondence to: Kathy K.H. Svoboda, Department of Biomedical Sciences, Texas A & M Health Science Center, BaylorCollege of Dentistry, 3302 Gaston Avenue, Dallas, TX 75246. E-mail: ksvoboda@bcd.tamhsc.edu

DOI 10.1002/dvdy.21637Published online 11 August 2008 in Wiley InterScience (www.interscience.wiley.com).

DEVELOPMENTAL DYNAMICS 237:2667–2675, 2008

© 2008 Wiley-Liss, Inc.

duced only by fibroblasts (Kallmanand Grobstein, 1965; Bernfield,1970).

Hay took the criticism as a per-sonal challenge to prove that epithe-lial cells could synthesize ECM mol-ecules, including type I collagen. Toaccomplish this she needed an epi-thelium that could be dissected freeof other tissue types, and would sur-vive sufficient lengths of time in cul-ture to demonstrate collagen synthe-sis and assembly. The salamanderlimb epithelium was not a good can-didate for in vitro studies becausethe tissue was not sterile and its bio-synthetic levels were insufficient toaddress the core questions. Hay readwith interest the paper published bythe Coulombres, which contained ex-quisite electron micrographs of colla-gen in the developing chick cornea.The micrographs showed that thefibrils were secreted and assembledin orthogonal arrays, that is, thefibrils were arranged in mat-like lay-ers, where within a layer the fibrilswere unidirectional. Any given layerwould have fibrils running at nearlya 90-degree angle with respect to thefibrils in the layer above it or belowit (Coulombre and Coulmbre, 1958).

EMBRYONIC CHICKCORNEAL EPITHELIALMODEL SYSTEM

Hay turned to the embryonic chickcorneal model system because the or-ganization of the matrix was similarto the pattern observed in the base-ment lamella of the salamander limb.The collagen fibrils first laid down bythe corneal epithelium were assem-bled into an acellular “primary”stroma in the absence of fibroblasts.This occurred 2 days before thestroma was invaded by neural crest-derived cells, which synthesized the“secondary,” mature stroma. Thesefeatures made the embryonic chickcornea a perfect model for her subse-quent experiments.

By the mid 1960s, Jean Paul Reveland Hay had completed a compre-hensive study of the fine structure ofthe developing chick cornea. Theirresults were published in a mono-graph that is referenced to this day(Hay and Revel, 1969). The early de-velopment of the cornea was de-

scribed, carefully documenting theattributes of the epithelial-derivedprimary stroma before fibroblast in-vasion. The fibrils made by the epi-thelium had a periodicity of approx-imately 600Å, with an interbandstriation pattern identical to fibrilsassembled in the secondary cornealstroma, synthesized by the fibro-blasts. At the time, this research

provided the most compelling evi-dence that epithelial cells producedthe collagenous primary stroma. Thediscovery of a second collagen, typeII (Miller and Matukas, 1969), wasstill 2 years away. If Hay’s embry-onic epithelial fibrillar collagen pa-pers had been published after thediscovery of type IV collagen, shewould still have been likely to face

Fig. 1. The culture system developed by Dodson, Meier, and Hay placing epithelium on killed lenscapsule. A: Epithelium cultured directly on lens capsule. B: Epithelium separated from lens capsuleby a Nucleopore filter barrier. Reprinted from Meier and Hay, 1975. C: Transmission electronmicrograph of a 24-hr culture showing epithelial cell processes extending into the filter. Reprintedfrom Hay and Meier, 1976. D: Corneal epithelia grown directly on one Nucleopore filter, 0.8 �m poresize (bottom curve), on stacks of 0.8�m pore size filters with lens capsule beneath the filter, ordirectly on lens capsule (top curve). Cultures were grown in the presence of 5�Ci/ml H3 proline andharvested at various times up to 24 hr. The bottom curve demonstrates the base line level ofcollagen synthesis, while the top curve corresponds to the “induced state.” Stimulation of syntheticactivity decreased with the increasing filter thickness, consistent with the observation that the cellprocesses traversed the filter contacting the lens capsule. Vertical bars � standard deviation forfour assays. Reprinted from Meier and Hay (1975).

2668 SVOBODA ET AL.

skepticism, because type IV wouldhave been considered “the” epithelialcollagen. Because her observationswere novel, they were bound to cre-ate discord. No matter. Hay’s persis-tence in her research, along withsubsequent work by others, has sup-ported her discovery that embryonicchick corneal epithelia synthesizefibrillar collagens. It is now knownthat this epithelium can secrete aminimum of three fibrillar collagens:type I, type II (Hendrix et al., 1982),and type V (Gordon et al., 1994).

INFLUENCE OF MATRIXON THE EMBRYONICCHICK CORNEALEPITHELIUM

The next series of experiments usingcultured corneal epithelia launched abody of work that has spanned 40years and remains an active field tothis day. James W. Dodson joinedHay’s laboratory as a postdoctoral fel-low in the late 1960s. He had grownchick skin in organ culture on a vari-ety of “cell-free” substrata (Dodson,

1967). These substrata were made bydissecting an organ known to have anabundant matrix and killing its cellswith low temperature, leaving the ma-trix intact. Dodson was quick to applythis approach to the corneal epithe-lium. By culturing sheets of cornealepithelia on “cell-free” lens capsules(Fig. 1A), he demonstrated that theepithelium produced a proline richECM, that was absent when the samecells were cultured on plastic. Thiswas additional evidence for collagensynthesis by epithelia (Dodson andHay, 1971). It also suggested to Haythat the components of the ECMmight regulate embryonic growth anddifferentiation. Her group exploited acombination of morphological and au-toradiographic approaches to examinethe response of corneal epithelia tovarious substrata (Hay and Dodson,1973; Dodson and Hay, 1974). Ste-phen Meier, a graduate of MichaelSolursh’s laboratory, joined the Haylab in 1972. His background in chon-drogenesis included cell–matrix inter-actions, and he brought biochemicalanalytical skills to the project. Hiswork provided the initial quantita-tion of collagen synthesis (Meier andHay, 1974b) and glycosaminoglycan(GAGs) synthesis (Meier and Hay,1973) induced in corneal epithelia inresponse to specific substrata. Fromthis work, it became evident that anappropriate substratum was requiredfor the embryonic epithelium to syn-thesize and assemble a primarystroma in vitro.

Meier next designed a set of exper-iments to demonstrate that physicalcontact with the substratum en-hanced the epithelial response. Heobserved that the basal surface ofcorneal epithelial sheets, enzymati-cally released from the 5-day chickembryo, contained protrusions termed“blebs.” Meier sandwiched Nucleo-pore filters between the epithelialsheet and killed lens capsule (Fig.1B,C). When a single filter was usedin the culture system (Meier andHay, 1975), the basal blebs tra-versed the filter’s pores and madecontact with the lens capsule sub-stratum (Fig. 1C). However, whenmore than one filter was interposed,the response to the lens capsule di-minished in proportion to the num-ber of filters (Fig. 1D). In addition,

Fig. 2. Diagrams and electron micrographs illustrating the effects of different experimental conditionson the organization of the basal cell surface. A: The basal surface blebbed when the basal lamina wasremoved by EDTA or enzyme treatment, and the blebbing persisted on Millipore filters in the presenceof nonmatrix proteins or glycosaminoglycans. B: Soluble collagens, fibronectin and laminin added tothe medium stimulated the bleb retraction and reformation of the basal actin cytoskeleton. Reprintedfrom Sugure and Hay (1982). C: Corneal epithelial tissues were detergent extracted and treated with S-1fragments of heavy meromyosin before EM preparation. C-1 is a typical bleb demonstrating a core ofactin filaments decorated with myosin S-1 fragments, which aligns on the actin filaments indicating thepolarity of the filament that is pointing toward the plasma membrane (indicated by small arrows) andsome apparently inserting into a dense plaque (dp). Intermediate filaments (IF) were in the basal cellarea. C-2 is a smaller bleb with some actin filaments parallel to the cell surface (arrowheads) and othersperpendicular to the surface. C-3 presents a cell 6 hr after immersion in soluble type IV collagen (100�g/ml). The actin bundle (MF) was in the basal compartment of this slightly tangential section containinga damaged plasma membrane (PM). The actin filaments were organized in opposite directions (smallarrows). Scale bars � 0.2�m. Reprinted from Sugrue and Hay (1982). D. Hay’s concept drawingillustrating the influence of the ECM on the actin cytoskeleton. It was used in lab meetings and titled“Hands across the membrane.”

EMBRYONIC CHICK CORNEAL EPITHELIAL MODEL 2669

there was a direct relationship be-tween the size of the pore and theresponse to the substrate (Meier andHay, 1975). This was strong evi-dence that cell contact with the sub-stratum was required for stimulat-ing the epithelial cells to synthesizecollagen. Not only did these experi-ments reinforce the concept thatthe developing corneal epitheliumneeded attachment to ECM to inducesynthesis of the primary stroma, butin their 1975 papers Meier and Hayproposed that epithelial cell pro-cesses might contain specific surfacesites for collagen-ECM interaction(Meier and Hay, 1975), foreshadow-ing the discovery of integrins.

MATRIX INDUCED ACTINREORGANIZATIONDuring the same time period, RichardHynes, while exploring the differencesbetween normal and tumor cells, iden-tified and isolated a large molecularweight protein that bound to normalcells (Hynes, 1973). Hynes termed theprotein LETS (large, external trans-formation-sensitive) protein. It waslater proven to be a ubiquitous extra-cellular protein, and renamed fi-bronectin (Wartiovaara et al., 1974;Mautner et al., 1977). In the latterpart of the 1970s, several laboratoriesdemonstrated that extracellular bun-

Fig. 3. Diagrams of several models depicting the possible relation of extracellular matrix (ECM)molecules to the cell surface. All models envision plasma membrane receptors for one or moremolecules. CO, collagen; FN fibronectin; PG, proteoglycans; HA, hyaluronic acid. A: Model basedon the data from Sugrue and Hay (1982). B: Model based on data from Singer (1979) and Kleinmanet al. (1981). C: Model of FN interaction with fibroblasts by Hynes et al. (1982). D: Model ofglycosaminoglycans and the mesenchymal cell surface from Toole (1981). Reprinted from Hay,(1981a,b).

TABLE 1. Summary of Stimulators and Inhibitors of Corneal Epithelial Actin Reorganizationa

No Effect on Actin ReorganizationStimulate Organization of the ActinCortical Mat

Blocked Actin CorticalMat Organization

Albumin Collagen (I–IV) #�2�1, �3�1, �6�1, �v�5, �6�4 Actin inhibitors (Cytochalasin D)Hyaluronic Acid Fibronectin - �3�1, �5�1, �6�1, �9�1, �v�1, �v�6, �6�4 Phospho-tyrosine inhibitorsAntibodies IgG Laminin (type I) �2�1, �3�1, �6�1, �v�1, �v�5, �6�4 Decreasing Rho protein or activityProteoglycans and

glycosaminoglycansincluding CS, HS, HSPG

Vitronectin - �v�3, �v�6 ROCK inhibitors

Serum LPA* PI-3 Kinase inhibitorsBombesin* MAP Kinase inhibitors

aMolecules that had no effect on organizing the corneal epithelial actin cortical mat included proteoglycans, glycosaminoglycans,serum, albumin and antibodies (Sugrue and Hay, 1982). Most of the extracellular matrix (ECM) molecules that bound to integrinreceptor subunits (Stepp, 2006) reorganized the actin cytoskeleton: these included collagens, fibronectin, laminin (Sugrue and Hay,1982) and vitronectin (Svoboda, unpublished). Molecules that directly stimulated the Rho pathway, including lysophosphatidic acid(LPA) and bombesin, reorganized the actin cortical mat faster (15 min) than integrin stimulation (2–6 hr; Svoboda et al., 1999b).Inhibitors of actin reorganization or the signaling pathways (broad phospho-tyrosine blockers, MAP-kinase, PI-3 kinase, ROCKinhibitors, or interference with Rho or its activity) blocked reorganization of the actin cortical mat (Svoboda and Hay, 1987;Svoboda, et al., 1999b, 2004, Chu et al., 2000; Reenstra et al., 2002).

#Integrin subunits identified in corneal epithelia (Stepp, 2006).*G-protein receptor- direct Rho stimulator with an Actin response in 15 min.

2670 SVOBODA ET AL.

dles of fibronectin, attached to thebasal surface of epithelial cells, werealigned with the submembranous ac-tin filaments in the cytoplasm (Hyneset al., 1978; Singer, 1979; Heggenesset al., 1978). Additional work showedthat fibronectin was able to induce therearrangement of actin filaments infibroblasts and myoblasts (Ali et al.,1978). The term “fibronexus” wascoined to describe this conjoined ex-tra- and intracellular structural com-plex (Singer, 1979).

Stephen Sugrue joined Hay’s lab inthe late 1970s to determine whetherpurified ECM proteins, such as spe-cific collagen types (I, II, and IV),laminin and fibronectin, would pro-mote the synthesis of differentiationmarkers by corneal epithelium. Sug-rue’s data demonstrated that epithe-lia grown on Millipore filters, with nosoluble matrix component added tothe culture, continued to bleb for theduration of the experiments (12–14hr; Fig. 2A). No fibrillar collagen wassynthesized. However, if soluble colla-gen (COLL), laminin (LAM), or fi-bronectin (FN) were added to the me-dium, blebs were withdrawn within2–6 hr and the basal cell surfaces flat-tened (Table 1; Fig. 2B). Sugrue andHay demonstrated that soluble ECMmolecules added to the medium boundto the basal epithelial surface, andthese induced and regulated collagensynthesis. In addition, it was shownthat the binding sites for the solublematrix components were protein-aceous (Sugrue and Hay, 1986). Incontrast, epithelia grown on filters inthe presence of added albumin, IgG, orglycosaminoglycans continued to formbasal blebs, and did not attain theflattened basal phenotype (Table 1;Fig. 2A). Also, epithelia cultured onsolid matrices, such as glass, contin-ued to bleb in the absence of ECMcomponents (Sugrue and Hay, 1981).It was clear that cell–ECM interac-tions were central to actin reorganiza-tion and expression of a differentiatedphenotype. Ultimately, myosin S-1decoration definitively establishedthat the basal submembranous fila-ment bundles were F-actin (Fig. 2C).Of interest, new protein synthesis wasnot required for the corneal epithelialcells to respond to collagen or laminin,but was needed for the cell’s responseto fibronectin (Sugrue and Hay, 1982).

These data suggested that actin reor-ganization in response to the ECMwas responsible for bleb retraction(Fig. 2B). The intracellular basalsheet occupied by actin was termedthe “actin cortical mat” (ACM). Hay,who was very artistic, drew her inter-pretation of the interactions betweenthe basal epithelia and the extracellu-lar matrix in an illustration referredto by lab members as “hands acrossthe membrane” (Fig. 2D). This draw-ing was used by the group for concep-tual discussions, but was never pub-lished. It was given to Sugrue when heleft Harvard because it represented asummary of his work. As a tribute toHay, he framed the original drawingand presented it back to her at a din-ner following the Matrix and Morpho-genesis conference, held in Boston to

honor her 75th birthday and her con-tributions to the field.

CELL–MATRIXINTERACTIONS

Because ECM binding to the cell sur-face clearly involves specific molecularinteractions, a search for ways toblock these interactions was begun inseveral labs. Once techniques were re-fined to purify ECM molecules, mono-clonal and polyclonal antibodies weregenerated against whole moleculesand their proteolytic fragments. Thegoal was to identify blocking antibod-ies that would inhibit cell adhesion tospecific substrates. In the early 1980s,review papers highlighted the experi-mental data demonstrating evidencefor integral membrane linkers (Fig. 3).

Fig. 4. Confocal microscopy of actin in corneal epithelial sheets. Schematic drawing of embryoniccorneal epithelia isolation and culture. A: Epithelia were isolated as a sheet of tissue and placed ona polycarbonate filter, basal side down, then cultured at the air–media interface, B: Epitheliaisolated without the basal lamina (�BL) extended basal cellular processes termed “blebs.”C: Sheets of corneal epithelia were stained with fluorescent phalloidin to label all filamentous actin(F-actin). In single confocal optical sections, the blebs are punctate spots in the plane of the basalcells. The cortical actin near cell membranes is also in the micrograph. D: If the tissue is culturedwith soluble extracellular matrix (ECM) molecules, the basal actin reorganizes into an actin corticalmat (ACM) that phalloidin labels as bundles of F-actin that align from cell to cell after COL treatmentfor 2 hr. Scale bar � 10 �m. Reprinted from Svoboda et al. (1999b).

EMBRYONIC CHICK CORNEAL EPITHELIAL MODEL 2671

Hay was an author on several of thesereviews (Hay, 1977, 1981b, 1984;Hynes et al., 1982; Singer, 1979). Yetthe identity of the transmembranelinker proteins proved elusive. Fi-nally, in the mid 1980s the first inte-grin was identified and cloned

(Tamkun et al., 1986); for review see(Hynes, 2004). As with many proteinfamilies, once the first integrin wascloned, the flood gates opened andmany others were quickly identified.Hynes organized a Gordon Conferencein early 1987, inviting everyone he

knew who was working on cell adhe-sion proteins. By the end of that week,the integrin family was pieced to-gether and a common terminology ac-cepted (Hynes, 1987). In the late1980s and early 1990s many groupswere using the newly characterizedintegrin antibodies to identify tissue-specific subunits. The Trinkaus-Ran-dall and Gipson labs were the first todescribe integrin subunits in the cor-neal epithelial cells (Stepp et al.,1990a,b; Trinkaus-Randall et al.,1990; Grushkin-Lerner and Trinkaus-Randall, 1991; Stepp, 2006).

In 1982, Kathy Svoboda joined theHay lab as a post doctoral fellow, be-ing interested in the role of cytoskel-etal reorganization during morpho-genesis. Her goal was to determinewhether actin reorganization in re-sponse to the ECM is a necessary pre-requisite for increased collagen pro-duction within cultured cornealepithelia. Her data indicated that in-creased collagen synthesis was, in-deed, actin dependent, and if the actinmat was disrupted, the rough endo-plasmic reticulum became displaced

Fig. 5.

Fig. 5. Confocal images of epithelia grown onfilters with fluorescently labeled matrix compo-nents. A,B: Corneal epithelia were isolatedwithout the basal lamina (�BL), incubated withmedia containing fluorescein isothiocyanate–fi-bronectin (FITC-FN) for short time intervals (15or 30 min.). The epithelia were rinsed, fixed andviewed on the confocal microscope in the xzoptical plane. The intensity of bound FITC-la-beled FN was recorded at the basal cell surfacein a nonuniform pattern. The intensity level ofFITC-FN increased from 0 to 30 min as ana-lyzed with NIH Image. C: Corneal epithelia �BLwere isolated, incubated with media containingFN or COL for a short time interval (0, 10, 15,20, or 30 min). The cultures were rinsed andimmersed in control non-ECM media, and thenincubated for a total time of 2 hr. Samples werefixed and stained with fluorescent phalloidin,and then groups coded (numbered) and scoredby at least two people as �, �/�, or �actincortical mat (ACM). The experiments were re-peated three times, with an n � 5–7 epithelia/group. Folded and damaged tissue was notscored. Greater than 80% of the epithelia re-sponded to the ECM molecules within 20–30min, 22% with 10 min. D: Tritium labeled colla-gen was added to cultured epithelium in thepresence of unlabeled collagen (1, 10, 50, 100or 500 �g/ml) for 30, 60 or 120 min. Increasedbinding of 3H-collagen was observed up to 100�g/ml but binding decreased at 500 �g/ml,suggesting competition at high levels of coldcollagen. Reprinted from Svoboda et al.(1999b).

2672 SVOBODA ET AL.

from the basal cytoplasm (Svobodaand Hay, 1987).

IMPACT OF HAY’SCORNEAL EPITHELIALRESEARCH ONEPITHELIAL CELLBIOLOGY

For the most part in the late 1980s,Hay’s work with the corneal epithelialsheet model was superseded by re-search on epithelial mesenchymaltransitions. However, her impact oncorneal epithelial cell biology contin-ues to this day. For example, afterleaving the Hay laboratory, Svobodahas continued to use the corneal epi-thelial sheets in her own laboratory,further characterizing binding of thecell sheets to soluble matrix, identify-ing the signal transduction pathways,and actin associated proteins involvedin the reorganization of the actin cor-tical mat (Svoboda, 1990, 1992;Khoory et al., 1993; Hirsch et al.,1994; Svoboda et al., 1999a). To visu-alize the changes that Hay’s grouphad been showing biochemically, con-focal microscopy was refined as a toolto evaluate actin reorganization (Fig.4). En face images with fluorescentlylabeled phalloidin showed bright fluo-rescent spots when the cells were op-tically sectioned transverse to the ep-ithelial blebs (Fig. 4C), and large actinbundles formed in parallel on the baseof the cells as the ACM was reorga-nized (Fig. 4D). This morphologicalmethod unraveled the three-dimen-sional organization of actin and its as-sociated proteins, intracellular or-ganelles and mRNAs for collagen Iand actin. (Svoboda, 1991, 1992;Khoory et al., 1993; Yeh and Svoboda,1993, 1994; Svoboda et al., 1999a).The time-course of binding was mea-sured using fluorescein isothiocyante(FITC) -labeled fibronectin and colla-gen I (Fig. 5A,B). The corneal epithe-lial sheets needed 15 min in the pres-ence of soluble matrix before the ACMbegan reforming, and the mat wascomplete by 2 hr (Fig. 5C). Bindingand competition experiments alsoshowed that binding was saturable(Fig. 5D). These experiments comple-mented both Meier and Sugrue’s ex-periments on the corneal epithelial re-

sponse to collagen (Meier and Hay,1974a; Sugrue and Hay, 1981, 1982).

Tyrosine phosphorylation of manysignaling proteins was observed incorneal epithelia as they bound exog-enous fibronectin or collagen, begin-ning within 15 min of exposure to thesoluble matrix (Svoboda et al., 1999b).Focal adhesion kinase (FAK) andp190RhoGAP were tyrosine phosphor-ylated within 5 min. Other MAP ki-nase signaling pathways (Erk-1, 2)and PI-3 kinase were activated later(30–60 min). These phosphorylationswere required for actin reorganiza-tion, as shown with a tyrosine and Srckinase inhibitor, Herbimycin A (Table1; Svoboda et al., 1999b). Further-more, blocking the MAP kinase andPI-3 kinase pathways with specific in-hibitors also prevented actin reorgani-zation and decreased ECM binding(Table 1) (Chu et al., 2000). Monolayercultures of corneal epithelial cellsgrown on laminin, fibronectin and col-lagen I showed that ACM reorganiza-tion was accompanied by an increasein intracellular pH (Wu et al., 1995).Reorganization of the actin corticalmat in the epithelial sheets was Rho-dependent (Reenstra et al., 2002) andROCK-dependent (Svoboda et al.,2004). All of this work has been de-rived from the elegantly simple con-clusion of Elizabeth Hay that epithe-lial cells in the embryonic chick corneacan synthesize and assemble fibrillarcollagen and do so in response to itsunique ECM.

Some of Hay’s other trainees havecontinued their research on the cor-nea. May Griffith, a former postdoc-toral fellow, has developed cornealconstructs, resulting in many publica-tions, including one in Science (Grif-fith et al., 1999). Others, includingStephen Sugrue, expanded their in-terest to other epithelia. In Sugrue’scase, this led to the discovery of themolecule pinin (Ouyang et al., 1997;Shi et al., 2000; Zimowska et al., 2003;Joo et al., 2005), a multifunctionalmolecule involved in cell–cell interac-tions and much more. Additionally,colleagues saw the potential of Hay’scorneal model system: Ilene Gipsonadapted the model to rabbit cornea(Geggel and Gipson, 1985; Gipson etal., 1985; Spurr and Gipson, 1985).This began a body of work unravelingthe components of the adhesion com-

plex between the epithelium and thestroma in the cornea.

PERSPECTIVES

This embryonic corneal epithelial or-gan culture model provides investiga-tors with a more physiological systemthan monolayer cell cultures. Epithe-lial–stromal interactions, actin-asso-ciated protein changes, the influenceof soluble matrix, and downstreamsignaling molecules, have all beenstudied using it. This model preservesmany characteristics of in vivo epithe-lium, including apical-basal polarityand cell–cell junctions. Corneal epi-thelial cells have a rich integrin sub-set that also responds to ECM mole-cules (Table 1, Stepp, 2006). Powerfulnew techniques are emerging usingfluorescently tagged proteins and re-al-time cell imaging that will enable amuch better understanding of the in-tracellular events that take placewhen the epithelium is challengedby trauma, toxins or transformingagents.

Because signaling triggers alter-ations in cell metabolism, in gene ex-pression and in cell behavior, contin-ued use of the corneal epithelial modelcould link signal transduction cell bi-ology with embryonic morphogenesis.

ACKNOWLEDGMENTSThis review is dedicated to the memoryof Elizabeth D. Hay. We thank SteveSugrue for helpful insights and for pro-viding the “hands across the mem-brane” illustration. We thank Develop-mental Biology and Journal of CellBiology for permission to reprint fig-ures. Dedication: K.K.S. and M.K.G.were funded by NIH/NEI.

REFERENCES

Ali IU, Hynes RO, Ali IU, Hynes RO. 1978.Effects of LETS glycoprotein on cell mo-tility. Cell 14:439–446.

Bernfield MR. 1970. Collagen synthesisduring epitheliomesenchymal interac-tions. Dev Biol 22:213–231.

Chu CL, Reenstra RW, Orlow DL, SvobodaKKH. 2000. Erk and PI3 kinase are nec-essary for collagen binding and actin reor-ganization in avian corneal epithelia. In-vest Ophthalmol Vis Sci 41:3374–3382.

Coulombre AJ, Coulmbre JL. 1958. Cor-neal development. II. Transparencychanges during rapid hydration. Am JOphthalmol 46:276–280; discussion 281.

EMBRYONIC CHICK CORNEAL EPITHELIAL MODEL 2673

Dodson JW. 1967. The differentiation ofepidermis. I. The interrelationship ofepidermis and dermis on embryonicchicken skin. J Embryol Exp Morphol17:83–105.

Dodson JW, Hay ED. 1971. Secretion ofcollagenous stroma by isolated epitheliagrown in vitro. Exp Cell Res 65:215–220.

Dodson JW, Hay ED. 1974. Secretion ofcollagen by corneal epithelium. II. Effectof the underlying substratum on secre-tion and polymerization of epithelialproducts. J Exp Zool 189:51–72.

Geggel HS, Gipson IK. 1985. Removal ofviable sheets of conjunctival epitheliumwith dispase II. Invest Ophthalmol VisSci 26:15–22.

Gipson IK, Friend J, Spurr SJ. 1985.Transplant of corneal epithelium to rab-bit corneal wounds in vivo. Invest Oph-thalmol Vis Sci 26:425–433.

Gordon MK, Foley JW, Birk DE, Fitch JM,Linsenmayer TF. 1994. Type V collagenand Bowman’s membrane. Quantitationof mRNA in corneal epithelium andstroma. J Biol Chem 269:24959–24966.

Griffith M, Osborne R, Munger R, Xiong X,Doillon CJ, Laycock NL, Hakim M, SongY, Watsky MA. 1999. Functional humancorneal equivalents constructed from celllines. Science 286:2169–2172.

Grushkin-Lerner L, Trinkaus-Randall V.1991. Localization of integrin and synde-can in vivo in a corneal epithelial abra-sion and keratectomy. Curr Eye Res 10:75–85.

Hay ED. 1977. Interaction between the cellsurface and extracellular matrix in cor-neal development. In: Lash JW, BurgerMM, editors. Cell and tissue interac-tions. New York: Raven Press. p 115–137.

Hay ED. 1981a. Cell biology of the extra-cellular matrix. 1st ed. New York: Ple-num Press. 468 p.

Hay ED. 1981b. Extracellular matrix.J Cell Biol 91:205s–223s.

Hay ED. 1984. Cell-matrix interaction inthe embryo: cell shape, cell surface, cellskeletons, and their role in differentia-tion. In: Hay ED, editor. Role of extracel-lular matrix in development. New York:Alan R. Liss, Inc. p 379–409.

Hay ED, Dodson JW. 1973. Secretion ofcollagen by corneal epithelium: I. Mor-phology of the collagenous products pro-duced by isolated epithelia grown on fro-zen-killed lens. J Cell Biol 57:190–213.

Hay ED, Revel JP. 1963. Autoradiographicstudies of the origin of the basement la-mella in Ambystoma. Dev Biol 7:152–168.

Hay ED, Meier S. 1976. Stimulation of cor-neal differentiation by interaction be-tween cell surface and extracellular ma-trix. II. Further studies on the natureand site of transfilter “induction”. DevBiol 52:141–157.

Hay ED, Revel JP. 1969. Fine structure ofthe developing avian cornea. Basel, Swit-zerland: Karger.

Heggeness MH, Ash JF, Singer SJ. 1978.Transmembrane linkage of fibronectin tointracellular actin-containing filaments

in cultured human fibroblasts. Ann N YAcad Sci 312:414–417.

Hendrix MJ, Hay ED, von der Mark K,Linsenmayer TF. 1982. Immunohisto-chemical localization of collagen types Iand II in the developing chick cornea andtibia by electron microscopy. Invest Oph-thalmol Vis Sci 22:359–375.

Hirsch MS, Law LY, Trinkaus-Randall V,Svoboda KKH. 1994. The intracellulardistribution of vinculin and alpha 2 inte-grin in epithelial cells and chondrocytes.Scanning 16:275–284.

Hynes RO. 1973. Alteration of cell-surfaceproteins by viral transformation and byproteolysis. Proc Natl Acad Sci U S A70:3170–3174.

Hynes RO. 1987. Integrins: a family of cellsurface receptors. Cell 48:549–554.

Hynes RO. 2004. The emergence of inte-grins: a personal and historical perspec-tive. Matrix Biol 23:333–340.

Hynes RO, Ali IU, Mautner VM, Destree A.1978. LETS glycoprotein: arrangementand function at the cell surface. BirthDefects Orig Artic Ser 14:139–153.

Hynes RO, Destree AT, Wagner DD. 1982.Relationships between microfilaments,cell-substratum adhesion, and fibronec-tin. Cold Spring Harb Symp Quant Biol46(pt 2):659–670.

Joo JH, Alpatov R, Munguba GC, JacksonMR, Hunt ME, Sugrue SP. 2005. Reduc-tion of Pnn by RNAi induces loss of cell-cell adhesion between human corneal ep-ithelial cells. Mol Vis 11:133–142.

Kallman F, Grobstein C. 1965. Source ofcollagen at epitheliomesenchymal inter-faces during inductive interaction. DevBiol 11:169–183.

Khoory W, Wu E, Svoboda KKH. 1993. In-tracellular relationship between actinand alpha-actinin in a whole corneal ep-ithelial tissue. J Cell Sci 106:703–717.

Kleinman HK, Klebe RJ, Martin GR,Kleinman HK, Klebe RJ, Martin GR.1981. Role of collagenous matrices in theadhesion and growth of cells. J Cell Biol88:473–485.

Mautner V, Hynes RO, Mautner V, HynesRO. 1977. Surface distribution of LETSprotein in relation to the cytoskeleton ofnormal and transformed cells. J Cell Biol75:743–768.

Meier S, Hay ED. 1973. Synthesis of sulfatedglycosaminoglycans by embryonic cornealepithelium. Dev Biol 35:318–331.

Meier S, Hay ED. 1974a. Control of cornealdifferentiation by extracellular materi-als. Collagen as a promoter and stabi-lizer of epithelial stroma production. DevBiol 38:249–270.

Meier S, Hay ED. 1974b. Stimulation ofextracellular matrix synthesis in the de-veloping cornea by glycosaminoglycans.Proc Natl Acad Sci U S A 71:2310–2313.

Meier S, Hay ED. 1975. Stimulation of cor-neal differentiation by interaction betweencell surface and extracellular matrix. I.Morphometric analysis of transfilter “in-duction”. J Cell Biol 66:275–291.

Miller EJ, Matukas VJ. 1969. Chick carti-lage collagen: a new type of alpha 1 chain

not present in bone or skin of the species.Proc Natl Acad Sci U S A 64:1264–1268.

Ouyang P, Zhen YY, Sugrue SP. 1997.Cloning and analysis of cDNA encodingmurine pinin. Gene 197:115–120.

Reenstra WR, Orlow DL, Svoboda KKH.2002. ECM-stimulated signaling and ac-tin reorganization in embryonic cornealepithelia are Rho dependent. InvestOphthalmol Vis Sci 43:3181–3189.

Shi Y, Tabesh M, Sugrue SP. 2000. Role ofcell adhesion-associated protein, pinin(DRS/memA), in corneal epithelial mi-gration. Invest Ophthalmol Vis Sci 41:1337–1345.

Singer II. 1979. The fibronexus: a trans-membrane association of fibronectin-containing fibers and bundles of 5 nmmicrofilaments in hamster and humanfibroblasts. Cell 16:675–685.

Spurr SJ, Gipson IK. 1985. Isolation of cor-neal epithelium with Dispase II orEDTA. Effects on the basement mem-brane zone. Invest Ophthalmol Vis Sci26:818–827.

Stepp MA. 2006. Corneal integrins andtheir functions. Exp Eye Res 83:3–15.

Stepp MA, Park CC, Spurr-Michaud S, Gi-pson I. 1990a. Identification of integrinsin stationary and migrating corneal epi-thelia. Invest Ophthalmol Vis Sci31:538a.

Stepp MA, Spurr-Michaud S, Tisdale A,Elwell J, Gipson IK. 1990b. Alpha 6 beta4 integrin heterodimer is a component ofhemidesmosomes. Proc Natl Acad SciU S A 87:8970–8974.

Sugrue SP, Hay ED. 1981. Response ofbasal epithelial cell surface and cytoskel-eton to solubilized extracellular matrixmolecules. J Cell Biol 91:45–54.

Sugrue SP, Hay ED. 1982. Interaction ofembryonic corneal epithelium with exog-enous collagen, laminin, and fibronectin:role of endogenous protein synthesis.Dev Biol 92:97–106.

Sugrue SP, Hay ED. 1986. The identifica-tion of extracellular matrix (ECM) bind-ing sites on the basal surface of embry-onic corneal epithelium and the effect ofECM binding on epithelial collagen pro-duction. J Cell Biol 102:1907–1916.

Svoboda KKH. 1990. Spatial localization ofcollagen mRNA and cytoskeletal pro-teins in embryonic corneal epitheliumwith confocal image analysis. Scanning90:112–113.

Svoboda KKH. 1991. Intracellular localiza-tion of types I and II collagen mRNA andendoplasmic reticulum in embryonic cor-neal epithelia. J Cell Sci 100:23–33.

Svoboda KKH. 1992. Embryonic cornealepithelial actin alters distribution in re-sponse to laminin. Invest OphthalmolVis Sci 33:324–333.

Svoboda KKH, Hay ED. 1987. Embryoniccorneal epithelial interaction with exog-enous laminin and basal lamina is F-actin dependent. Dev Biol 123:455–469.

Svoboda KKH, Orlow DL, Ashrafzadeh A,Jirawuthiworavong G. 1999a. Zyxin andvinculin distribution at the cell-extracel-lular matrix attachment complex

2674 SVOBODA ET AL.

(CMAX) in corneal epithelial tissue areactin dependent. Anat Rec 254:336–347.

Svoboda KKH, Orlow DL, Chu CL, Reen-stra WR. 1999b. ECM-stimulated actinbundle formation in embryonic cornealepithelia is tyrosine phosphorylation de-pendent. Anat Rec 254:348–359.

Svoboda KKH, Moessner P, Field T, Ace-vedo J. 2004. ROCK inhibitor (Y27632)increases apoptosis and disrupts the ac-tin cortical mat in embryonic avian cor-neal epithelium. Dev Dyn 229:579–590.

Tamkun JW, DeSimone DW, Fonda D, Pa-tel RS, Buck C, Horwitz AF, Hynes RO.1986. Structure of integrin, a glycopro-tein involved in the transmembranelinkage between fibronectin and actin.Cell 46:271–282.

Toole BP. 1981. Glycosamionglycans inmorphogenesis. In: Hay ED, editor. Cellbiology of the extracellular matrix. NewYork: Plenum Press. p 259–294.

Trinkaus-Randall V, Newton AW, FranzblauC. 1990. The synthesis and role of integrinin corneal epithelial cells in culture. InvestOphthalmol Vis Sci 31:440–447.

Wartiovaara J, Linder E, Ruoslahti E, Va-heri A. 1974. Distribution of fibroblastsurface antigen: association with fibril-lar structures of normal cells and lossupon viral transformation. J Exp Med140:1522–1533.

Wu XY, Svoboda KKH, Trinkaus-RandallV. 1995. Intracellular distribution of cy-toskeletal and adhesion proteins in re-

sponse to corneal substrata. Exp Eye Res60:445–458.

Yeh B, Svoboda KKH. 1993. Intracellularlabeling of �-actin mRNA using reversetranscriptase incorporated biotin dUTPinto actin cortical mat of corneal epithe-lial cells. Micron 24:595–602.

Yeh B, Svoboda KKH. 1994. Intracellulardistribution of beta-actin mRNA is polar-ized in embryonic corneal epithelia.J Cell Sci 107:105–115.

Zimowska G, Shi J, Munguba G, Jack-son MR, Alpatov R, Simmons MN, Shi Y,Sugrue SP. 2003. Pinin/DRS/memA in-teracts with SRp75, SRm300 andSRrp130 in corneal epithelial cells. In-vest Ophthalmol Vis Sci 44:4715–4723.

EMBRYONIC CHICK CORNEAL EPITHELIAL MODEL 2675