Role of SH2-Containing Tyrosine Phosphatase Shp2 in Mouse Corneal Epithelial Stratification

Abstract

Purpose.: Shp2 protein tyrosine phosphatase mediates a wide variety of receptor tyrosine kinases (RTK) cell signaling. Herein, we investigate the role of Shp2 in corneal morphogenesis and homeostasis.

Methods.: Shp2 was conditionally knocked out (Shp2^cko ) in Krt14-rtTA;tet-O-Cre;Shp2^f/f triple transgenic mice administrated with doxycycline (Dox) from postnatal day 1 (P1) to P10, P15, and P25, respectively. In addition, corneal epithelial debridement was performed in adult (P42) mice treated with or without Dox for 8 days (from P42–P50). Mouse eyes were then subjected to histology and immunohistochemistry.

Results.: Shp2^cko revealed impaired stratification of conjunctival and corneal epithelia during morphogenesis. Likewise, Shp2^cko failed to restore epithelial stratification after a corneal epithelial wound in adult Shp2^cko . At the cellular level, the ratio of proliferating cell nuclear antigen (PCNA-positive)/total basal cells remained unchanged, but cells in G2/M (survivin-positive) phase was significantly increased in Shp2^cko as compared with those in the control littermate. Interestingly, deltaN-p63 (ΔNp63) expression and the asymmetric division of the basal cells were coincidentally dampened in Shp2^cko . Transmission electron microscopic study showed that desmosome and hemidesmosome densities were reduced in the corneal epithelium of Shp2^cko . Immunohistochemistry also demonstrated that expression of E-cadherin/β-catenin junction and laminin-β1 was extensively downregulated in Shp2^cko . On the other hand, corneal epithelium lacking Shp2 remained positive for K14, Pax-6, and keratin 12 (K12), suggesting that Shp2 was dispensable for the corneal epithelial-type differentiation.

Conclusions.: These data argued that Shp2 deficiency predominantly impacted p63-dependent cell division and cell adhesive ability, which resulted in the impairment of stratification during corneal epithelial development and wound healing.

Introduction

The stratification of epithelial cells forms a multilayer that serves as a primary barrier to prevent the invasion of pathogens and antigens from external environment, such as skin and ocular surface, where the stratification of epithelium can avoid epithelial cell defect duo to continuous exfoliation of superficial cells. ¹ However, the cellular and molecular mechanisms by which epithelial stratification is initiated during development and maintained have not been clearly identified. Previous studies have found that Shp2, a member of the protein tyrosine phosphatase family is ubiquitously expressed and plays a fundamental role in many cell signal pathways. ^2–5 One of these is the Ras/MAPK pathway, which impacts several facets of cellular behaviors, such as proliferation, apoptosis, and migration. ^6–8 All of these would influence cell stratification. 

Shp2 naturally remains in a folded conformation as its two N-terminal SH2 domains bind to the central protein tyrosine phosphatase domain, keeping the phosphatase in an auto-inhibited state. Shp2 is activated in the presence of phosphotyrosine residues of the epidermal growth factor (EGF) receptors when extracellular ligands bind. When Shp2 binds to the phospho-EGF receptor (active form), other scaffolding proteins, such as Grb2 and Gab1, are able to aggregate and continue signal transduction down the pathway. ^9–11 Because Shp2 interacts directly with activated EGFR, these findings suggest that the presence of the tyrosine phosphatase is essential to the Ras/MAPK pathway and maybe, therefore, necessary for the development and maintenance of proper epithelial stratification. 

Using the corneal epithelium as a model, the present study aims to identify the role of Shp2 in corneal epithelial stratification in the corneal morphogenesis and homeostasis during wound healing. Two strains of transgenic mice, K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} and K14-rtTA;tetO-Cre;Shp2^flox/+ , with the latter as the control, were employed in this study. Our results revealed that Shp2 was required for the initiation of corneal epithelial stratification, which occurred just before eyelid opening in the second week of postnatal age (P) in rodents. ¹ Shp2 was also indispensable for the restratification of epithelium during epithelial debridement wound healing. A possible mode of action of Shp2 in these processes is also discussed. 

Materials and Methods

Mice

Compound transgenic mouse strain K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} was generated through the natural mating of single transgenic mouse lines K14-rtTA,¹² tetO-Cre, ¹³ and Shp2^{flo xed} , ¹⁴ as seen in Figure 1A. K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} transgenic males were mated with K14-rtTA;tetO-Cre;Shp2^{flo x/+} females. The resulting mouse littermates were induced at P0 since previous studies have shown that ablation of Shp2 during the early embryonic stage results in an open eyelid defect at birth, ¹⁴ which could introduce other factors that might impact the proper formation of the corneal epithelium. Therefore, to ensure that the data was solely attributed to the gene ablation, the mice were induced at P1, and their corneas were collected at P10, P15, and P25 for immunostaining. All mice were maintained in mixed genetic background. Experimental animals were housed under pathogen-free conditions in accordance with institutional guidelines. Animal care and use conformed to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. All animal protocols were approved by the Institutional Animal Care and Use Committee of the University of Cincinnati. 

Figure 1

 

K14⁺ cell type–specific and Dox-inducible ablation of Shp2 gene in the transgenic mouse. (A) Schematic presentation showing Dox-dependent ablation of Shp2 gene alleles. Double transgenic mouse, K14-rtTA;TC, serves as epidermal-specific Dox-inducible driver. Shp2 floxed allele (Shp2^f/f ) mouse line serves as Dox-inducible effector. Intercross between K14-rtTA;TC and Shp2^f/f generated triple transgenic mice K14-rtTA;TC;Shp2^f/f . Administration of Dox resulted in the deletion of exone4 of the Shp2 gene only in K14-positive basal cells in all stratified epithelial tissues including skin and ocular surface epithelia. (B) K14-rtTA;TC;Shp2^f/+ (control) and K14-rtTA;TC;Shp2^f/f (Shp2^cko ) mouse littermates administered with Dox in the dam via mother's milk from P0 to P10, P15, and P25, respectively. It should be noticed that mice lacking Shp2 alleles exhibited wavy fur at P15 and P25 as compared with the control littermates. This partially recapitulated phenotype in Waved-1 (TGF-α deficient) and Waved-2 (EGFR deficient) mouse strains. (C) Hematoxylin and eosin histologic analysis of corneal stratification corresponding to (B). (D) Immunohistochemistry showed that immunoreactivity of Shp2 protein was completely undetectable only in the corneal epithelium, but remained strong and positive in the corneal endothelium of the Dox-treated K14-rtTA;TC;Shp2^f/f mouse (compare right panels to left panels). Abbreviations: ce, corneal epithelium; co, cornea; cs, corneal stroma; cn, corneal endothelium. Scale bar: 100 μm in (D).

Administration of Doxycycline (Dox)

To induce Shp2 gene ablation in adult animals, K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} mice (P42) were injected once intraperitoneally (IP) with Dox (80 μg/g body weight; Clontech Laboratories, Mountain View, CA) dissolved in 1× PBS (pH 7.4) at a concentration of 10 mg/mL and fed Dox-chow (1 g/kg chow; Bioserv, Frenchtown, NJ) ¹⁵ ad libitum. Control animals were injected with PBS and fed regular chow. Neonates (P1) were administrated with Dox by feeding the nursing mother with Dox chow ad libitum. 

Histology Analysis

Mouse samples were fixed overnight in 4% paraformaldehyde (PFA) in PBS, followed by dehydration in a series of ascending alcohol and paraffin embedding. Deparaffinized sections (5 μm) were stained with hematoxylin and eosin (H&E) or periodic acid Schiff reagent (PAS). 

Immunofluorescence and Immunohistochemistry

Tissue sections (5 μm) were deparaffinized, rehydrated, and subjected to antigen retrieval in sodium citrate buffer (10 mM sodium citrate, 0.05% Tween, pH 6.0) at boiling temperature for 30 minutes. Eyelid sections were then blocked with 3% BSA in PBS containing 0.025% Nonidet P40 for 1 hour at room temperature, then incubated overnight at 4°C with the primary antibodies diluted in the same buffer. After three washes with PBST (PBS, 0.1% Tween 20), slides were incubated at room temperature for 1 hour with AlexaFluor-488- or AlexaFluor-555-conjugated secondary antibodies (Invitrogen, Inc., Carlsbad, CA) and 1 μg/mL 4′,6-diamidino-2-phenylindole (DAPI) as a nuclear counterstain, washed with PBST again, and mounted with Mowiol (Sanofi-Aventis, Paris, France). Sections were examined and photographed using a Zeiss microscope equipped with a camera (Axiocam Mrm; Carl Zeiss, Thornwood, NY). For data acquisition, we used the Axiovision 4.6 software (Carl Zeiss). The following antibodies were used: rabbit anti–β-Catenin antibody (#C2206, Sigma-Aldrich, St. Louis, MO); rabbit anti–keratin 12 (K12) antibody ¹⁶ ; rabbit anti-K14 (PRB-155P; Covance, Princeton, NJ), rabbit anti-K18 (clone E431-1, cat# 04-586, EMD; Millipore Corporation, Billerica, MA); rabbit anti-Shp2 (#3752; Cell Signaling Technology, Danvers, MA), rabbit anti-Survivin (clone 71G4B7; Cell Signaling Technology); mouse anti-p63 monoclonal antibody (4A4; Neomarkers, Fremont, CA); mouse anti–proliferating cell nuclear antigen (PCNA, PC10:sc-56; Santa Cruz Biotechnology, Inc., Santa Cruz, CA); anti–E-cadherin (clone 67A4; Chemicon, Inc., Temecula, CA), goat anti–laminin-β1 (sc-6018; Santa Cruz Biotechnology, Inc.), Secondary Alexa488-, Alexa555-labeled antibodies were purchased from Molecular Probes, Inc. (Eugene, OR). 

Detection of Apoptosis In Situ

Cell apoptosis status in Dox-treated and nontreated groups were subjected to TUNEL assay using Click-iT TUNEL Assay kit (Invitrogen, Inc.) according to the manufacturer's protocol. 

Transmission Electron Microscopy

Cornea samples were obtained and fixed in 0.1M cacodylate buffer (pH7.4) containing 3% glutaraldehyde and 2% paraformaldehyde for 2 hours at 4°C and then were preserved in 0.1 M cacodylate buffer (pH 7.4) containing 0.5% glutaraldehyde at 4°C overnight. After refixation in 1% osmium tetraoxide (OsO4) for 1 hour at 4°C, cornea samples were washed in 0.1 M cacodylate buffer (pH 7.4) three times for 10 minutes, and then dehydrated in a graded ethanol series and embedded in Epon 812 epoxy resin (Polysciences, Inc., Warrington, PA). Ultrathin 50-nm sections were stained with uranyl acetate and lead citrate and images were photographed with a Hitachi 7500 Transmission Electron Microscope (Hitachi, Tokyo, Japan) equipped with AMT Digital camera. 

Wound Healing Experiment

A 2-mm diameter central corneal epithelial debridement was generated in the right eye using Algerbrush II Rust Ring Remover (Alger Equipment Co., Inc., Lago Vista, TX) and mice were then separated into two groups. One group of mice was fed with Dox (+Dox) chow and another group of mice was fed with regular chow (−Dox) for 8 days. 

Statistical Analysis

A two-tailed Student's t-test (Excel; Microsoft, Redmond, WA) was used in the analysis of the percentage of PCNA-, deltaN-p63 (ΔNp63)-, and survivin-positive cells. All quantification data are presented as mean ± SEM. Student's t-test was used to analyze the significance of difference; P less than 0.05 was considered statistically significant. 

Results

K14+ Cell Type–Specific and Dox-Inducible Ablation of Shp2 Gene in the Transgenic Mouse

Conventional Shp2 knockout mice died early in the embryonic development. To examine the function of Shp2 in ocular surface epithelial morphogenesis, we generated a K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} triple–transgenic mouse strain, in which conditional Shp2 ablation was induced in basal cells of the stratified epithelia upon Dox treatment (Fig. 1A). We first administered Dox from embryonic day 0.5 (E0.5) to P0 (at birth) and noticed that, unlike K14-rtTA;tetO-Cre;Shp2^{flo x/+} (control) littermate with closed eyelid, K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} (Shp2^cko ) displayed an open eyelid phenotype (Supplementary Fig. S1) and died within 24 hours due to defects in esophagus and upper airway (unpublished data). To avoid neonatal death and to allow the investigation of corneal epithelium stratification occurring just prior to eyelid opening at P12 to P14 in the mouse, ^1,17,18 we administered mouse neonates with Dox from P1 via nursing mother fed Dox–chow as described in Materials and Methods. With this protocol of Dox treatment, Shp2^cko mice could survive and remained healthy during the weaning stage (up to P25), but gradually became moribund after a couple of days of weaning. This was also due to obstruction of the esophagus, which inhibited food uptake (data not shown). Therefore, we examined Shp2^cko mice and their control littermates at P10, P15, and P25, respectively. We did not find a significant difference in general appearance between control and Shp2^cko at P10. However, wavy fur was noticed in Shp2^cko at P15 and P25, which partially resembled the phenotypes of TGF-α (Wave-1) and EGFR (Wave-2) deficient mice ^19–21 (Fig. 1B). This observation suggested that Shp2 was likely to mediate TGF-α/EGFR signaling pathway in epidermal and ocular surface tissues. 

Shp2 Deficiency Caused Failure of Ocular Surface Epithelial Stratification

Histologic analysis revealed that corneal epithelium consisting of two cell layers at P9 stratified to four to five cell layers at P15 and to six to seven cell layers at P25 in the control mice. In contrast, corneal epithelium failed to do so and remained at one to two cell layer(s) in Shp2^cko at P15 and P25 (Fig. 1C; Supplementary Fig. S2). Immunohistochemical staining with anti-Shp2 antibody confirmed that the Shp2 protein was totally undetectable in the corneal epithelium, but remained strong and positive in the corneal endothelium of the Shp2^cko mouse at P25. (Fig. 1D). In addition to the corneal epithelium, other ocular surface epithelia, including eyelid mucocutaneous and conjunctival epithelium, also failed to stratify into multilayers. Moreover, the ectoderm-derived meibomian gland and goblet cells were malformed in the Shp2^cko ocular surface epithelia (Supplementary Figs. S3, S4). In this paper, we focused on the stratification of the corneal epithelium. 

Shp2 Deficiency Lost Polarized Basal Cells and Reduced the Density of Desmosomes Between Cells and Hemidesmosomes Between Cell-Matrix

Next, we took a closer look regarding the ultrastructure of corneal epithelium under a transmission electron microscope as shown in Figure 2. The corneal epithelium exhibited seven to eight cell layers in the control mice. In contrast, the corneal epithelium of Shp2^cko mice only displayed two cell layers at P25 (Fig. 2A versus Fig. 2B). The corneal epithelium of Shp2^cko mice lost the distinctive polymorphic epithelium from basal to superficial layers and only flattened cells were exhibited in the cornea of the Shp2^cko mice (Fig. 2B). Moreover, the density of desmosomes between epithelial cells and hemidesmosomes between epithelial cells and the basement membrane (BM) were obviously reduced in Shp2^cko mutants as compared with the control littermates (Fig. 2B′′ versus Fig. 2A′′′). However, no early phase of apoptosis with condensed or peripheralized chromatin was found. Likewise, no apparent autophagosome membranes showed in either control or Shp2^cko corneal epithelia. These data suggested that Shp2 deficiency might impact cell polarity, cell–cell and cell-matrix adhesion, but little cell death (please see below). 

Figure 2

 

Shp2 deficiency reduced the density of desmosome between cells and hemidesomesone between cell-matrix. Transmission electronic micrographs of control group (A, A′, A′′, A′′′) and Shp2^cko (B, B′, B′′). Control corneal epithelium exhibited seven to eight cell layers at P25 (A), but displayed only one to two cell layers in Shp2^cko (B). Unlike control epithelium, which showed distinct cell morphology of different layers (i.e., basal columnar cells), suprabasal wing cells, and superficial flattened cells, Shp2 deficient epithelium exhibited only flattened cells. It should be noticed that the density of desmosome (open triangles) between cells and hemidesmosome (arrowheads) between cell and basement membrane were dramatically reduced in Shp2 deficient corneal epithelium. No early phase of apoptosis with condensed or peripheralized chromatin was found. Likewise, no apparent autophagosome membranes showed in the Shp2 deficient corneal epithelium. Abbreviations: Nu, nucleus; epi, epithelium.

Role of Shp2 in Proliferation and Division Orientation of Corneal Epithelial Basal Cells

Given that epithelial stratification has been at least in part conferred by the balanced symmetric cell division (SCD) and asymmetric cell division (ACD) of the basal cells, ^22,23 we performed immunofluorescence staining to detect PCNA and ΔNp63 to examine whether impaired stratification in Shp2^cko mice was attributed to defects in epithelial cell proliferation and/or division orientation. We found that 75.6 ± 4.5% and 78.8 ± 3.1% of the total basal cells in the control and Shp2^cko corneal epithelia, respectively, were PCNA-positive (Figs. 3A–C). Notably, some cells right above their basal layer were also PCNA-positive in the control group, but little of such were found in the Shp2^cko cornea (compare insets of Figs. 3A′ and 3B′). Given that ΔNp63 is required for the cell cycle progression, ACD, and epidermal stratification, ^24–26 as we expected, ΔNp63-positive cells were significantly reduced from 90.9 ± 5.2% in the control cornea down to 45.2 ± 2.4% in those of the Shp2^cko (Figs. 3D–F). We then performed immunohistochemistry to examine the expression of survivin, which is only expressed in the G2/M phase and localizes to the mitotic spindle by interaction with tubulin during mitosis. ²⁷ We found that the percentage of survivin-positive cells/total basal cells increased from 22.2 ± 2.6% in the control cornea up to 31.3 ± 2.6% in those of the Shp2^cko (Figs. 3G–I). More interestingly, most of the basal cells in corneal epithelium of control underwent perpendicular cell division, while those in the Shp2^cko turned into a parallel orientation relative to the BM (compare Figs. 3G and 3H). These data strongly suggested that Shp2 deficiency not only affected cell cycle progression through G2/M phase, but also dramatically reduced ACD, leading to the impaired stratification. 

Figure 3

 

Role of Shp2 in cell proliferation and division of the corneal epithelial basal cells. Immunofluorescent stained micrographs of PCNA (A, A′, B, B′), ΔNp63 (D, D′, E, E′), and survivin (G, H) in control (A, A′, D, D′, G) and Shp2^cko (B, B′, E, E′, H) mice at P25. Noted that Shp2 deficiency did not impact cell proliferative activity (compare B, B′ with A, A′), but downregulated ΔNp63 (compare D, D′ with C, C′). Histogram of quantitative analyses represents the percentage of cells positive for PCNA (C) and ΔNp63 (F) in the corneal epithelial basal layer. The percentage of survivin-positive (G2/M of the cell cycle) basal cells in Shp2^cko was significantly increased as compared with that of control cornea (G, H, I). More interestingly, basal cells of the control cornea underwent division perpendicular to basement membrane (dotted lines) to move up, while cell division orientation in Shp2^cko is parallel to basement membrane (I). Scale bars: 50 μm. *P < 0.05.

Role of Shp2 in Cornea-Type Differentiation

We next investigated whether corneal epithelial differentiation was altered by Shp2 ablation. As shown by the immunofluorescence staining, Pax-6^28–29 (a transcription factor essential for ocular surface epithelial differentiation) and K12, the marker of cornea-type epithelium differentiation, remained strongly positive in the nonstratified epithelium of the Shp2^cko cornea (Figs. 4B, 4B′, 4D, 4D′). Interestingly, keratin 14 (Krt14) expressions, which was positive for all basal cells in the control cornea (Figs. 4E, 4E′) was interrupted along the corneal epithelium (Figs. 4F, 4F′) in the cornea of Shp2^cko mice. These data illustrated that Shp2 deficiency did not impact cornea-type differentiation but downregulated the expression of K14, an epithelial progenitor cell marker. 

Figure 4

 

Shp2 deficiency did not impact cornea-type differentiation. Immunofluorescent-stained micrographs of Pax-6 (A, A′, B, B′), K12 (C, C′, D, D′), and K14 (E, E′, F, F′) in control (A, A′, C, C′, E, E′) and Shp2^cko (B, B′, D, D′, F, F′) mice at P25. Noted that Shp2 deficiency did not impact Pax-6 or K12 (markers for corneal epithelium), but downregulated the expression of K14, an epithelial progenitor cell marker. Scale bars: 50 < μm.

Shp2 Deficiency Altered Cell–Cell and Cell-Matrix Adhesion

Because of the downregulation of desmosomes and hemidesmosomes observed in corneal epithelium of the Shp2^cko mice, we examined cell adhesive functions. Immunofluorescence staining unambiguously demonstrated that the expression of E-cadherin, a calcium-dependent cell–cell adhesion molecule ³⁰ (Figs. 5B, 5B′) and β-catenin (Figs. 5D, 5D′) are dramatically downregulated in Shp2^cko . Furthermore, laminin-β1, ³¹ an extracellular matrix (ECM) component of the epithelium BM was completely diminished in corneal epithelium of Shp2^cko (Figs. 5F, 5F′). These data argued that Shp2 deficiency might compromise cell–cell and cell-BM adhesive function. Taken together, Shp2 function played a pivotal role in the epithelium stratification during normal corneal morphogenesis. 

Figure 5

 

Shp2 deficiency altered cell–cell and cell-matrix adhesion. Immunofluorescent-stained micrographs of E-cadherin (A, A′, B, B′), b-catenin (C, C′, D, D′), and laminin-β1 (E, E′, F, F′) in control (A, A′, C, C′, E, E′) and Shp2^cko (B, B′, D, D′, F, F′) mice administered with Dox from P1 to P25. It should be noticed that Shp2 deficiency drastically downregulated the expression of cell adhesive molecules E-cadherin/β-catenin and basement membrane component laminin-β1 in corneal epithelium. Scale bars: 50 < μm.

Shp2 Is Required for the Restoration of Stratification After Corneal Epithelial Debridement

To address the necessity of Shp2 for maintaining corneal epithelial homeostasis, K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} triple–transgenic mice were subject to central corneal epithelial abrasion and then were treated with Dox (Shp2^cko ) or with regular chow (control) for 8 days as shown in Figure 6A. In a naïve cornea, Cre recombinase effectively knocked out Shp2 alleles from the basal epithelial cells upon Dox treatment (Supplementary Fig. S5C, arrows). This resulted in the loss of basal lamellar structure (compare Figs. 6D and 6B, arrowheads), but did not crushed down stratified epithelium. In wounded eye, the initial re-epithelialization to cover the 2-mm diameter wound area behaved no differently between +Dox and −Dox as judged by the corneal epithelial fluorescein staining at 24 hours after wound healing (data not shown). This result suggested that Shp2 was not required for cell migration during early phase wound closure. Therefore, we did not attempt to compare the migrating leading edge within 6 to 24 hours after the wound. However, when examining if the epithelium could restore seven to eight layers and display typical basal lamellar structure 8 days after epithelial abrasion during wound healing, we noticed that these only happened when Shp2 remained intact in the absence of Dox (Fig. 6C, arrowheads). In contrast, epithelial restratification was compromised (Fig. 6E) when Shp2 was ablated in the presence of Dox (Supplementary Fig. S5C). Concomitantly, laminin-β1 expression was drastically reduced in the unwounded eye (compare Supplementary Figs. S5G and S5E) and completely diminished during wound healing (Supplementary Fig. S5H) when Shp2 was ablated, while laminin-β1 expression was upregulated during epithelial restratification (Supplementary Fig. S5F) when Shp2 was expressed (Supplementary Fig. S5B). Moreover, as shown in Supplementary Figure 6E, 80.0 ± 5.3% (+Dox) versus 82.7 ± 3.5% (−Dox) of basal cells in the unwounded eye and 76.3 ± 3.4% (+Dox) versus 78.5 ± 2.1% (−Dox) of basal cells in the wounded eye were PCNA-positive, suggesting that proliferative activity was not altered when Shp2 was ablated in the unwounded cornea (compare Supplementary Figs. S6C and S6A) and in corneal restratification during wound healing (compare Supplementary Figs. S6D and S6B). On the other hand, as revealed in Figure 8E, the percentage of ΔNp63-positive cells was reduced approximately 3.3-fold (from 63.9 ± 4.4% to 19.4 ± 4.1%) when Shp2 was ablated in the unwounded cornea (compare Supplementary Figs. S6C′ and S6A′). More interestingly, we found that nearly half (48.7 ± 4.8%) of the re-epithelialized epithelial cells expressed ΔNp63 (Supplementary Fig. S6B′), but only 3.6 ± 2.5% of those were ΔNp63-positive when Shp2 was ablated during wound healing (Supplementary Fig. S6D′). Taken together, these data argued that Shp2 deficiency caused downregulation of ΔNp63 and the impairment of stratification during corneal epithelial wound healing. 

Figure 6

 

Shp2 is required for the restoration of stratification after corneal epithelial debridement. Adult (P42) K14-rtTA;tetO-Cre;Shp2^f/f mice were administered with or without Dox (80 μg/g body weight) via intraperitoneal injection. A 2-mm diameter central corneal epithelial debridement was generated in right eye (OD) with Algerbrush II Rust Ring Remover. Left eyes (OS) were wounded. Mice were then separated into two groups; group I mice were fed with regular chow (−Dox) and group II mice were fed with Dox chow (+Dox) for 8 days (A). Histologic analysis was performed by H&E staining (B–E). Noted that +Dox predominantly affected basal lamellar formation, albeit no significant changed of epithelium thickness in unwounded cornea (comparing arrowheads shown in [D] with [B]). Epithelium stratification was restored in wounded cornea in which basal lamellar formation was also seen in −Dox (C). In contrast, only two to three cell layers were formed without basal lamellar formation in +Dox cornea (E).

Discussion

The stratified, nonkeratinized corneal epithelium serves as a barrier that contributes to maintain corneal transparency and rigidity, which warrant normal visual function. In mice, the anatomic cornea demarcation from conjunctiva and the expression of cornea-type specific keratin 12 coincide with eyelid closure at E14.5. ³² At this stage, the primitive corneal epithelium consists of two cell layers and remains constant until the time of eyelid opening in P11 to P12. Interestingly, just prior to eyelid opening, corneal epithelium quickly stratify from two cell layers to up to six to seven cell layers at P25 and maintains as such throughout their lifetime ¹ (Fig. 1). Similarly, experimental corneal epithelial wound healing recapitulates, at least in part, the developmental process of corneal epithelial cell stratification by re-epithelializing the denuded surface with two cell layers and then stratifying to restore the full thickness of a healthy corneal epithelium ³³ (Fig. 6). Therefore, epithelial stratification in corneal morphogenesis and homeostasis may be achieved by the similar mechanism. 

Growth factors including the EGF family, keratinocyte growth factor, hepatocyte growth factor, insulin-like growth factor (IGF) insulin, TGF-β, and so on, have been shown to involve in corneal epithelial development and wound healing, ³⁴ but the key intracellular molecule that transmits signaling to control epithelial stratification has not been documented. In this study, our data argued that Shp2 might serve as a hub for various growth factors to relay the cellular signaling needed for cornea morphogenesis and homeostasis. The mutant strain in which Shp2 was conditionally knocked out in K14-positive basal cells exhibited wavy fur that resembled a TGF-α KO phenotype. As for ocular surface tissues, TGF-α knockout animals displayed uniformly thinner corneal epithelia than normal mice. ¹⁹ No stratification took place throughout the ocular surface epithelium of Shp2^cko mice and defects in their derived glands and goblet cells were observed. This suggested that in addition to TGF-α, the aforementioned growth factors might combine efforts to achieve ocular surface epithelial morphogenesis. Perhaps Shp2 directly integrates and relays signaling from these growth factors down to the nucleus for gene regulation and tissue morphogenesis. It should be noted that unlike TGF-α KO corneas, ²⁰ which revealed profound inflammation, little inflammation was found in those of Shp2^cko mice. Therefore, corneal inflammation seen in TGF-α KO corneas ²⁰ might not be triggered by the epithelium but stroma or other cell types. The phenotypes presented in this study were not likely attributed to the inflammation, but to the autonomous effect of Shp2 deficiency. 

The processes of establishing a stratified corneal epithelium have been termed “XYZ hypothesis”—X, the proliferation of basal epithelial cells; Y, the contribution to the cell mass by centripetal movement of peripheral cells; and Z, the epithelial cell loss from the surface. ³⁵ The underlying molecular genetics of each component of this hypothesis has never been examined; however, we found that Shp2^cko failed to develop stratified, nonkeratinized ocular surface epithelia during development and to restore stratification after a corneal epithelium debridement wound. In line with the hypothesis, the X component in Shp2^cko cornea remained as active as a control cornea (Fig. 3C). Likewise, the Z component was probably unaffected, since we did not detect any cell death from Shp2^cko cornea (data not shown). However, the Y component was significantly influenced as the total epithelial basal cell count was dramatically reduced in Shp2^cko cornea, which might be due to the decrease of ΔNp63-positive cells from Shp2 deficiency (Fig. 3F). It has been shown that ΔNp63 isoforms promoted cell cycle progression through a dominant negative effect on p53³⁶ as well as transcriptional repression of p21 (CIP1/WAF1). ³⁷ Therefore, downregulation of ΔNp63 resulting from Shp2 deficiency perhaps slowed down or arrested cell cycle progression, leading to the overall cell mass reduction of the corneal epithelium. This statement is in agreement with the most recent published paper by Günschmann et al. ³⁸ in which conditional deletion of insulin/IGF-1 signaling (IIS) resulted in a mitotic checkpoint arrest and impaired epidermal stratification through FoxO-mediated p63 inhibition. Moreover, since p63 is required for proper cell division orientation and indeed plays an essential role for ACD of the epithelial basal cells during epithelial stratification and morphogenesis, ^25,39 reduction of p63 in Shp2^cko might result in a biased loss of ACD, leading to impaired stratification. Given that Shp2 can relay receptor tyrosine kinases (RTK) signaling, including IIS from membrane to nucleus, ⁴⁰ therefore, it makes sense that Shp2^cko manifests a similar phenotype to that of the loss of IIS and p63 gene, respectively, in corneal stratification. 

In epidermal tissues, p63 is required for proliferation, stem/progenitor cells regeneration, differentiation, and apoptosis during development. Embryonic stratification failed to initiate at E12.5 in mice lacking p63 and led to defects in limb, craniofacial, and all stratified epithelia and their derivatives such as mammary, lacrimal, and salivary glands. ⁴¹ These data demonstrated that p63 was an ectodermal-specific gatekeeper of epidermal fate during development. ^42,43 Ocular surface epithelia differentiation occurred at E14.5, but stratification began at approximately P12, ⁴⁴ therefore, the interpretation to studies concerning the role of p63 in the developing corneal epithelium at E12.5 to P0 should attribute to the cell fate determination and differentiation, but not to the cell dividing event. In our study, Shp2 deficiency was induced postnatally before stratification in differentiated ocular surface epithelia. We concluded that corneal epithelial stratification largely relied on p63-dependent ACD but not on cell differentiation as the mono/bilayer epithelium remained positive for Pax-6, K12, and K14 in Shp2^cko mice (Fig. 4). On the other hand, it has been documented that ΔNp63 isoforms upregulate cell adhesion molecules, increasing cellular adhesion in various epithelial cells. ^45–47 We observed a downregulation of desmosome and hemidesmosome as well as downregulation of E-cadherin, β-catenin, and laminin-β1 expression in Shp2 deficient corneal epithelium. This suggested that in addition to affecting cell division orientation, Shp2^cko compromised cell–cell and cell-BM adhesion and led to the impairment of corneal epithelium stratification. 

Previous studies have documented the three typical phases of corneal epithelial wound healing that occur after epithelial debridement: (1) cell migration, (2) cell proliferation, and (3) cell differentiation to restore corneal homeostasis and barrier function. ³³ Immediately following epithelial abrasion, re-epithelialization is initiated by the adjacent K12⁺ superficial cell sheet, which migrates downwards to cover the wound area. Since the superficial cells had functioning Shp2, migration and re-epithelialization were fine. However, the subsequent transient amplifying (TA) cell sheet (K14⁺) migration underneath the K12⁺ sheet was impaired due to the downregulation of p63 and the deficiency of hemidesmosomes, desmosomes, and the BM ECMs such as laminin-β1. We concluded that Shp2 function was to lay down a good foundation and relatively tight anchors for the basal cell to form a stratified epithelium. Nevertheless, it remains unknown (1) how Shp2 regulates desmosome components and ECMs of BM? Is it a transcriptional event and/or cytoskeletal event? (2) What is the downstream signaling event of Shp2? Is this mediated through Ras-Mek-Erk? ¹⁰ We are currently crossing K14-rtTA;TC;Shp2^f/f into Rosa^Kras(G12D) (constitutively active) ⁴⁸ to see if the phenotype can be rescued and to what extent. If so, Shp2-Ras-Mek-Erk axis plays an essential role in corneal epithelial stratification in vivo; otherwise, other signaling molecules such as focal adhesion kinase, nuclear factor of activated T cell, signal transducer and activator of transcriptions may be involved in the epithelial stratification. The feasibility of K14-rtTA;tetO-Cre;Shp2^{flo x/flo x} mouse strain in conjunction with other Dox-inducible transgenic and/or gene-targeted floxed alleles will help elucidate the molecular mechanism(s) underlying corneal epithelium stratification and ocular surface morphogenesis. 

Supplementary Materials

pdf S1 

Acknowledgments

The authors thank Scheffer C.G. Tseng (Tissue Tech, Inc., Ocular Surface Center, and Ocular Surface Research and Education Foundation, Miami, Florida) for his critical reading and comments of this manuscript. They also thank Hui-Chun Kung, Ya-Ling Chen, and Shih-Hsin Hsiao for the preparation of transmission electronic microscopic studies at the Microscope Center of Chang-Gung Memorial Hospital, Linko, Taiwan. 

Supported by National Institutes of Health/National Eye Institute Grants RO1 EY21501 (CYL) and EY13755 (WWK), Research to Prevent Blindness, Ohio Lions Foundation for Eye Research, National Science Council (Taiwan) Grant 1012314B182A056MY3, and Grants CMRPG3A1292 and BMRP868 from Chang-Gung Memorial Hospital, Linko (LKY). 

Disclosure: G.Y. Ng, None; L.-K. Yeh, None; Y. Zhang, None; H. Liu, None; G.-S. Feng, None; W.W.-Y. Kao, None; C.-Y. Liu, None 

Footnotes

 GYN and L-KY are joint first authors.