January 2025
Volume 14, Issue 1
Open Access
Cornea & External Disease  |   January 2025
Establishment of an Ex Vivo Human Corneal Endothelium Wound Model
Author Affiliations & Notes
  • Meng-Chen Tsai
    UCL Institute of Ophthalmology, University College London, London, UK
  • Alvena Kureshi
    Centre for 3D Models of Health and Disease, Division of Surgery and Interventional Science, University College London, London, UK
  • Julie T. Daniels
    UCL Institute of Ophthalmology, University College London, London, UK
  • Correspondence: Meng-Chen Tsai, University College London, 3/33, UCL Institute of Ophthalmology, 11-43 Bath Street, London EC1V 9EL, UK. e-mail: [email protected]
Translational Vision Science & Technology January 2025, Vol.14, 24. doi:https://doi.org/10.1167/tvst.14.1.24
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Meng-Chen Tsai, Alvena Kureshi, Julie T. Daniels; Establishment of an Ex Vivo Human Corneal Endothelium Wound Model. Trans. Vis. Sci. Tech. 2025;14(1):24. https://doi.org/10.1167/tvst.14.1.24.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: A human model able to simulate the manifestation of corneal endothelium decompensation could be advantageous for wound healing and future cell therapy assessment. The study aimed to establish an ex vivo human cornea endothelium wound model where endothelium function can be evaluated by measuring corneal thickness changes.

Methods: The human cornea was maintained in an artificial anterior chamber, with a continuous culture medium infusion system designed to sustain corneal endothelium and epithelium simultaneously. The corneal thickness was used to assess corneal endothelial cell function. Immunostaining was used to evaluate cell viability and endothelial cell marker expression, ZO-1 and Na/K ATPase.

Results: Human corneas with intact corneal endothelium were maintained in the ex vivo model for 28 days, showing normal corneal thickness with a clear and transparent appearance. Corneal endothelial cells were alive and expressed ZO-1 and Na/K ATPase at the end of the organ culture. The endothelium wounded corneas showed persistent corneal edema with an increase in corneal thickness at 654.6 ± 31.7 µm. Staining results showed that no cells migrated to cover the wound and no expression of ZO-1 and Na/K ATPase on the posterior surface of the cornea was found.

Conclusions: This study provided a novel method to establish an ex vivo human cornea organ culture model, where corneal endothelium function can be evaluated by accessing the corneal thickness.

Translational Relevance: The ex vivo model established in this study can provide an alternative to the animal model in studying corneal endothelium decompensation.

Introduction
Corneal clarity relies on well-functioning and sufficient cell density of endothelial cells to control the hydration of the cornea. Endothelium transplantation using donor tissue is the standard surgical approach for treating endothelium deficiency caused by disease or injuries.1 More efforts have been made toward a cell therapy approach to circumvent the donor tissue shortage and extend the donor supply for multi-recipients. In cell therapy, endothelial cells are expanded in vitro and transplanted in vivo using a cell carrier or directly injecting cells into the anterior chamber to restore endothelium function.2 
To test whether cell therapy can resume endothelial function by pumping water out of the cornea to restore corneal clarity, it is necessary to establish a wound model for the experiment. Many corneal endothelium-wound models have been proposed, including in vitro, in vivo, and ex vivo. The in vitro endothelium wound models are established by scraping a linear wound into a confluent cell monolayer in a tissue culture plate.3 This can be suitable for evaluating endothelial cell proliferation and migration and to see if cells could cover the wound area,4 but it might not be useful for evaluating the endothelial cell pump function. An alternative way to test endothelial pump function in vitro is by measuring the cell potential differences in an Ussing chamber.57 The Ussing chamber technique is used to measure the transepithelial or endothelial electrical resistance related to the integrity of the cellular barriers under investigation.8 Endothelial pump function can be evaluated in the Ussing chamber by using ouabain.9 However, this method still cannot directly assess the pump function in relation to reducing corneal thickness.6 On the other hand, time and special techniques are required to set up a complex Ussing chamber system, whereas the tissue can only survive in the Ussing chamber for hours, which might not be suitable for long-term cell function observation.10 
The in vivo endothelium models use live animals such as rabbits, nonhuman primates, and rats for experiments. The rabbit is the most prevalent animal species used for endothelial wound models. Corneal thickness and clarity changes can be evaluated; however, rabbit corneal endothelial cells possess regenerative and proliferative capacity, which human corneal endothelial cells lack.11 Therefore using rabbits to establish corneal endothelial models might not be able to recapitulate the biology and immunology of the human cornea. Rats have relatively low costs compared to other animals. However, considering the eye size and the endothelium regeneration ability, it might be difficult to assess the corneal thickness changes in cell therapy studies. Nonhuman primate eyes show similar characteristics to human eyes with limited in vivo endothelium regenerative capacity, whereas using nonhuman primates for research requires special regulation and high expenses.12 In addition, the tectonic structure of the animal cornea differs from that of the human cornea in terms of corneal curvature, diameter, thickness, and mechanical properties.1315 These are also essential parameters that could affect cell attachment in cell therapy transplantation. 
The ex vivo endothelium wound model is an alternative to the in vivo animal model, in which native animal or human cornea tissue is either used for testing the feasibility of cell therapy delivery methods16,17 or establishing an organ culture system for cell attachment and cell morphology observation.18,19 However, most of the published ex vivo models did not construct the eyeball's anterior segment structure to rebuild the anterior and posterior surfaces of the cornea. Assessing the endothelial cell function is challenging as the whole cornea is submerged in one culture medium.20,21 Therefore this study aimed to establish a human cornea organ culture model by mounting a cornea in an artificial anterior chamber (to simulate the anterior segment structure) and where the corneal endothelial cell function could be evaluated by assessing the corneal thickness changes. 
Methods
Human Cornea Preparation
Human corneal tissue was obtained from the National Health Service Blood and Transplant with consent for research studies and under ethics approval (10/H0106/57-2011ETR10). Corneas unsuitable for clinical transplantation were discarded and transferred for research use. The blood tests of all the donors were negative for HIV, HTLV, syphilis, HBV, HCV, and HEV. Human corneas were briefly washed with Dulbecco's phosphate-buffered saline (DPBS; Thermo Fisher Scientific, Waltham, MA, USA) containing 1% antibiotic-antimycotic (AA; Thermo Fisher Scientific) for ex vivo model experiments. The age of the donors was 76.5 ± 4.76 years old (Table). 
Table.
 
Summary of Donor Information
Table.
 
Summary of Donor Information
Establishment of a Human Cornea Ex Vivo Anterior Chamber Infusion Model
Three individual human corneas from different donors (n = 3 biological replicates) were used to establish a human cornea ex vivo anterior chamber infusion model. The corneas with intact Descemet membrane (DM) and endothelium were assigned as normal cornea. Corneas were maintained in an ex vivo anterior chamber infusion model for 28 days. The corneal endothelial cell function was evaluated by measuring the thickness changes at a designated time during the organ culture. At the end of the experiment, corneas were evaluated using a live/dead assay and flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase (Fig. 1). 
Figure 1.
 
The diagram illustrates the experimental design of establishing a human cornea ex vivo anterior chamber infusion model. Three individual human corneas were used to establish a human cornea ex vivo anterior chamber infusion model. N = 3 biological replicates. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay and flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase.
Figure 1.
 
The diagram illustrates the experimental design of establishing a human cornea ex vivo anterior chamber infusion model. Three individual human corneas were used to establish a human cornea ex vivo anterior chamber infusion model. N = 3 biological replicates. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay and flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase.
Three-Dimensional Printing of Medium Container Ring
To maintain the corneal epithelium in the ex vivo model, a suitable epithelium culture medium container is needed. A three-dimensional (3D) medium container ring was designed using online software (Tinkercad), and 3D-printed in a Form 3 printer (Formlabs, Somerville, MA, USA) using clear, biocompatible, United States Pharmacopeia class VI certified medical grade resin (Formlabs). The container ring consisted of a small ring on the bottom and a larger ring on the top, which can be attached to the top of the anterior chamber to contain epithelial culture medium for maintaining corneal epithelial cell growth. The 3D printing was completed and ready to be used after being washed with 99% isopropyl alcohol for 30 minutes and cured at 60°C for 60 minutes. The container ring was disinfected in 70% isopropyl alcohol for five minutes and left under UV light for one hour before use. 
Ex Vivo Anterior Chamber Infusion Model
Cornea tissue was briefly washed with DPBS before being mounted onto the platform of the artificial anterior chamber (Barron Precision Instruments, MI, USA). The cornea was positioned in the center of the platform and locked with a tissue retainer to create a closed chamber for culture medium infusion. There were two ports on the base of the chamber, allowing the culture medium to flow in and out. The inlet and outlet of the ports were connected to a peristaltic pump (100 µL/min), and a culture medium reservoir bottle was connected to the artificial anterior chamber to complete the flow circuit. The peristaltic pump was used to create a flow so the culture medium could flow back to the reservoir bottle. Also, a 0.2 µm filter was attached to the reservoir bottle to enable air exchange and equalize internal and external pressures. The upper medium container contained the epithelium culture medium (1:1 ratio of Dulbecco's modified Eagle's medium (Thermo Fisher Scientific) high glucose: F12, 10% fetal bovine serum (Thermo Fisher Scientific), and 1% AA). The reservoir bottle contained the endothelium culture medium (1:1 ratio of culture medium 199: F12, 5% fetal bovine serum, and 1% AA) (Fig. 2). The ex vivo anterior chamber infusion model was placed on a 3D orbital rotating shaker in the incubator at 37°C, with 5% CO2 in the air. The culture medium was changed on each other day (Figs. 3A, 3B). 
Figure 2.
 
Schematic representation of the ex vivo anterior chamber infusion model. This infusion model comprises a reservoir bottle, a peristaltic pump, and an artificial anterior chamber. The anterior chamber infusion system uses a peristaltic pump to create a flow so the culture medium can flow back to the culture medium reservoir bottle. The arrowhead indicates the direction of the culture medium flow. Cornea tissue is mounted on the anterior chamber and locked with a tissue retainer to create a closed chamber for culture medium infusion. The inlet and outlet of the ports were connected to a peristaltic pump (100 µL/min) and a culture medium reservoir bottle with connecting tubing to complete the flow circuit. A 0.2 µm filter was attached to the reservoir bottle to enable air exchange.
Figure 2.
 
Schematic representation of the ex vivo anterior chamber infusion model. This infusion model comprises a reservoir bottle, a peristaltic pump, and an artificial anterior chamber. The anterior chamber infusion system uses a peristaltic pump to create a flow so the culture medium can flow back to the culture medium reservoir bottle. The arrowhead indicates the direction of the culture medium flow. Cornea tissue is mounted on the anterior chamber and locked with a tissue retainer to create a closed chamber for culture medium infusion. The inlet and outlet of the ports were connected to a peristaltic pump (100 µL/min) and a culture medium reservoir bottle with connecting tubing to complete the flow circuit. A 0.2 µm filter was attached to the reservoir bottle to enable air exchange.
Figure 3.
 
Pictures of the ex vivo anterior chamber infusion model. (A) The Baron artificial anterior chamber with 3D-printed medium container ring. (B) The ex vivo anterior chamber infusion model was connected to a peristaltic pump and placed on a 3D orbital rotating shake in the incubator at 37°C, with 5% CO2 in the air.
Figure 3.
 
Pictures of the ex vivo anterior chamber infusion model. (A) The Baron artificial anterior chamber with 3D-printed medium container ring. (B) The ex vivo anterior chamber infusion model was connected to a peristaltic pump and placed on a 3D orbital rotating shake in the incubator at 37°C, with 5% CO2 in the air.
Cornea Thickness Measurement
The central corneal thickness was measured using an ultrasonic handheld pachymeter (P-1 pachymetry; Takagi Seiko Co., Ltd., Nagano-ken, Japan) at a speed of sound at 1640 m/s. Five applications were carried out for each cornea, and the individual measurements were recorded. The corneal thickness was measured at zero hours, two hours, four hours, six hours, and 12 hours at the beginning of the first day. After that, the corneal thickness measurements were carried out every 24 hours until the end of the experiment. 
Live/Dead Cell Viability Staining of the Flat-Mounted Cornea
Corneas were cut into four or six pie-shaped wedges for preparation. The corneas were washed with DPBS three times and incubated with DPBS containing Live/Dead viability kit reagent (2 µM Calcein AM and 4 µM ethidium homodimer-1 (Thermo Fisher Scientific) for 30 minutes at room temperature. The corneas were placed on glass slides covered with DPBS (Thermo Fisher Scientific) and then gently flattened using a rectangular coverslip. Fluorescent cell images were captured within 1 hour using ZEN Blue software connected to a LSM 700 confocal microscope (Zeiss, Oberkochen, Germany) operating at wavelengths of 526 nm for Calcein AM and 613 nm for ethidium homodimer-1. 
Flat-Mounted Immunofluorescence Staining
Corneas were cut into four or six pie-shaped wedges and rinsed with DPBS before staining. The corneas were washed with DPBS before being fixed in 10% neutral buffered formalin solution (Sigma-Aldrich Corp., St. Louis, MO, USA) for 30 minutes at room temperature and permeabilized in 0.25% Triton X-100 (Sigma-Aldrich Corp) in DPBS for five minutes at room temperature. The nonspecific binding sites were then blocked in DPBS containing 5% heat-inactivated goat serum (Thermo Fisher Scientific) for 30 minutes at 37°C. The primary mouse anti-ZO-1 antibody (339100; Thermo Fisher Scientific) was diluted in DPBS at 1:200 and incubated at 4°C overnight. The primary mouse anti-Na/K ATPase antibody (M7-PB-E9; Thermo Fisher Scientific) was diluted in DPBS at 1:100 and incubated at 4°C overnight. The secondary antibody Alexa Fluor 488 goat anti-mouse (Thermo Fisher Scientific) was diluted at 1:500 in DPBS and incubated with the cells for one hour at 37°C. F-actin was stained with Alexa Fluor 594 phalloidin (A12381;Thermo Fisher Scientific) diluted in DPBS at 1:400 and incubated for 30 minutes at room temperature. Cell nuclei were finally counterstained with Hoechst 33342 (Thermo Fisher Scientific) diluted at 1:1000 in DPBS for 15 minutes at room temperature. Three rinses in DPBS were performed between all steps, except the steps between blocking non-specific protein binding sites and incubating with primary antibody. The corneas were mounted with slow fade glass mounting medium (Thermo Fisher Scientific) and covered with a coverslip. Fluorescent images were captured using a Zeiss LSM 700 confocal microscope equipped with ZEN Blue software. 
Cryosection Staining
The corneas were rinsed with DPBS and fixed in 4% paraformaldehyde at 4°C overnight. Then, the sample was cryoprotected in 30% sucrose at 4°C overnight before being embedded in optimal cutting temperature compound (OCT; VWR, UK) and transferred on dry ice until the OCT became frozen. The sample containing OCT block was stored at −80°C until further use. Embedded sample blocks were serially sectioned at 7 µm using a Leica CM1850 cryostat (Leica) and transferred onto SuperFrost Plus adhesion microscope slides (Thermo Fisher Scientific). The slides were air dried at room temperature for at least 1 hour and then stored at −80°C. Section cutting followed the immunofluorescence staining method. Corneas were labeled for collagen IV using 10 µg/mL of anti-human collagen IV monoclonal antibody, Alexa Fluor 647 (1042, eBioscience; Thermo Fisher Scientific) and incubated at 4°C overnight. 
Establishing an Ex Vivo Corneal Endothelium Wound Model
Three pairs of human corneas from different donors were divided into two groups: normal corneas and endothelium wound corneas (n = 3 biological replicates). The endothelium wound cornea was established by peeling the central 8.5 mm DM and endothelium under a dissecting microscope to mimic the clinical manifestation of endothelium decompensation. The cornea was placed in an artificial anterior chamber platform with endothelium face up. An 8.5 mm trephine was used to make a circular marker on the posterior surface of the cornea. The central 8.5 mm DM and endothelium were manually peeled. Trypan blue was used to indicate the wound area after the DM and endothelium were removed. 
Corneas were maintained in the ex vivo anterior infusion chamber for 28 days. The corneal endothelial cell function was evaluated by measuring the thickness changes at designated times during organ culture. At the end of the experiment, corneas were processed for live/dead assay, flat-mounted immunofluorescence staining of ZO-1, Na/K ATPase, and F-actin, cryosection staining of collagen IV and hematoxylin and eosin (H&E) staining (Fig. 4). 
Figure 4.
 
The diagram illustrates the experimental design of establishing an ex vivo corneal endothelium wound model. Three pairs of human corneas (n = 3 biological replicates) were divided into two groups: normal corneas and endothelium wound corneas. The central 8.5 mm DM and endothelium were peeled to create an endothelium wound. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay, flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase, and cross-sectional staining.
Figure 4.
 
The diagram illustrates the experimental design of establishing an ex vivo corneal endothelium wound model. Three pairs of human corneas (n = 3 biological replicates) were divided into two groups: normal corneas and endothelium wound corneas. The central 8.5 mm DM and endothelium were peeled to create an endothelium wound. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay, flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase, and cross-sectional staining.
Statistical Analysis
All data were presented as mean ± standard error of the mean. Statistical analysis was performed using GraphPad Prism version 9.0.0 for macOS (GraphPad Software, San Diego, CA, USA) with one-way or two-way analysis of variance and Turkey multiple comparisons test. Differences between groups were considered statistically significant at P < 0.05. 
Results
The Human Cornea Thickness Changes in Ex Vivo Anterior Chamber Infusion Model
Three human corneas from three different donors with intact corneal endothelium were cultured and maintained in the ex vivo anterior chamber infusion model for 28 days (n = 3 biological replicates). At the beginning of the organ culture, the cornea thickness was 934.73 ± 52.76 µm, and then the thickness gradually decreased to a normal physiological level at 571.20 ± 38.05 µm within one day. After that, the thickness was maintained at a stable level for 28 days, from 549.60 ± 20.97 µm on day 2 to 570.53 ± 23.28 µm on day 28 (Fig. 5). 
Figure 5.
 
The human corneal thickness of ex vivo anterior chamber infusion model. Human corneas with intact corneal endothelium, normal group, were cultured and maintained in our models for 28 days. The initial thickness was 934.73 ± 52.76 µm, and the number decreased to the normal physiological level of 571.20 ± 38.05 µm within one day. Then, the corneas maintained a thickness of 570.53 ± 23.28 µm until day 28. Cornea thickness was measured by ultrasonic handheld pachymeter, with five measurements for each cornea. N = 3 biological replicates. Donor age and gender were 72 years old male, 74 years old female, and 75 years old male.
Figure 5.
 
The human corneal thickness of ex vivo anterior chamber infusion model. Human corneas with intact corneal endothelium, normal group, were cultured and maintained in our models for 28 days. The initial thickness was 934.73 ± 52.76 µm, and the number decreased to the normal physiological level of 571.20 ± 38.05 µm within one day. Then, the corneas maintained a thickness of 570.53 ± 23.28 µm until day 28. Cornea thickness was measured by ultrasonic handheld pachymeter, with five measurements for each cornea. N = 3 biological replicates. Donor age and gender were 72 years old male, 74 years old female, and 75 years old male.
Camera images showed the cornea was clear and transparent after 28 days of organ culture. Fluorescein staining indicated that the cornea epithelium was still intact, and only a few punctate epithelial defects were observed (Fig. 6). 
Figure 6.
 
Image of the human cornea after 28 days of ex vivo organ culture. (A) The human cornea remained clear and transparent on day 28. (B) Small punctate epithelial defects were observed after staining the cornea with 2% fluorescein.
Figure 6.
 
Image of the human cornea after 28 days of ex vivo organ culture. (A) The human cornea remained clear and transparent on day 28. (B) Small punctate epithelial defects were observed after staining the cornea with 2% fluorescein.
Small Numbers of Corneal Endothelial Cells Died After Ex Vivo Organ Culture
At the end of the organ culture experiment, human corneas were stained with Live/Dead assay to visualize the corneal endothelial cell viability. Most human corneal endothelial cells were still alive after 28 days of organ culture. However, some parts of the endothelial cells were missing, with small numbers of dead cells on the edge of the defect. Large endothelium defects were also observed, with large numbers of cells missing and displaying a moth-eaten appearance on the edge of the defect (Fig. 7). 
Figure 7.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture. Confocal microscopy (representative of n = 3 biological replicates) images showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Most of the cells were still alive after 28 days of organ culture. However, some endothelial cells were missing, with small numbers of dead cells on the edge of the defect (arrowhead). Some large defects were also observed, displaying a moth-eaten appearance. Scale bar: 50 µm.
Figure 7.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture. Confocal microscopy (representative of n = 3 biological replicates) images showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Most of the cells were still alive after 28 days of organ culture. However, some endothelial cells were missing, with small numbers of dead cells on the edge of the defect (arrowhead). Some large defects were also observed, displaying a moth-eaten appearance. Scale bar: 50 µm.
Corneal Endothelial Cells Retained ZO-1 and Na/K ATPase Expression After 28 Days of Ex Vivo Organ Culture
The corneal endothelial barrier plays a vital role in maintaining the cornea's transparency and the ideal hydration level of the stroma. Na/K ATPase and tight junction ZO-1 are the major components of this pump-leak barrier. After 28 days of organ culture, corneas were removed from the model for immunofluorescence staining of ZO-1 and Na/K ATPase. Labeling ZO-1 gave a green fluorescence staining of corneal endothelial cells with a clear hexagonal intercellular junction on the cell-cell border. Enlarged and polygonal corneal endothelial cells were observed after 28 days of organ culture. Most endothelial cells retained close cell-cell contact, whereas some were missing and left a small defect area behind. Transmembrane protein, Na/K ATPase, was also expressed on corneal endothelial cells after 28 days of organ culture, displaying a clear cell outline on the cell-cell border. Like ZO-1 staining, some cells were missing and lost cell hexagonality in Na/K ATPase staining (Fig. 8). 
Figure 8.
 
Immunofluorescence staining of ZO-1 and Na/K ATPase on the flat-mounted cornea. (A, C) ZO-1 was labelled with green fluorescence and displayed a clear intercellular outline on the cell-cell borders. Some cells became enlarged, with loss of cell hexagonality, and some cells were missing, leaving behind a small area of defect. (B, D) Na/K ATPase was labelled with green fluorescence and displayed a clear outline of cell membranes. Like ZO-1 staining, some endothelial cells were lost, leaving a small area of cell defect behind. Nuclei were counterstained with Hoechst 33342. Representative images, n = 3 biological replicates. Scale bar: 50 µm.
Figure 8.
 
Immunofluorescence staining of ZO-1 and Na/K ATPase on the flat-mounted cornea. (A, C) ZO-1 was labelled with green fluorescence and displayed a clear intercellular outline on the cell-cell borders. Some cells became enlarged, with loss of cell hexagonality, and some cells were missing, leaving behind a small area of defect. (B, D) Na/K ATPase was labelled with green fluorescence and displayed a clear outline of cell membranes. Like ZO-1 staining, some endothelial cells were lost, leaving a small area of cell defect behind. Nuclei were counterstained with Hoechst 33342. Representative images, n = 3 biological replicates. Scale bar: 50 µm.
Endothelium Wound Corneas Showed Persistent Edema and Thickened Cornea Compared to Normal Corneas
Three pairs of human corneas from three different donors (n = 3 biological replicates) were divided into two groups: normal cornea with intact DM and endothelium wound group with central 8.5 mm DM removal. The initial corneal thickness for normal corneas and wounded corneas were 934.67 ± 39.0 µm and 932.0 ± 39.1 µm, respectively. There is no significant difference between the normal and wounded groups (P = 0.69) at the beginning of organ culture. After four hours of organ culture, the thickness of the normal cornea decreased to 676.27 ± 41.8 µm, which was significantly lower than the thickness of the wounded cornea, 739.2 ± 46.3 µm (P < 0.05). After that, the corneal thickness of normal corneas was maintained at a stable level for 28 days, from 558.93 ± 30.5 µm on day 2 to 577.20 ± 19.3 µm on day 28. On the other hand, the corneal thickness of the wounded cornea was significantly thicker than the corneal thickness of normal corneas, from 701.73 ± 54.7 µm on day 2 to 654.6 ± 31.7 µm on day 28. Overall, the mean corneal thickness difference between the normal and wounded groups was 112.4 ± 6.9 µm, statistically significant (P < 0.05; Fig. 9A). Camera images also showed that the cornea was clear and transparent at the end of organ culture, while the wounded cornea showed an irregular light reflection of the corneal surface with epithelium edema (Figs. 9B–E). 
Figure 9.
 
The normal and wounded corneal thickness in an ex vivo anterior chamber infusion model. (A) Three pairs of human corneas were divided into two groups: normal cornea and wounded cornea. Endothelium wounded cornea was made by removing the central 8.5 mm of DM and corneal endothelium. The thickness of wounded corneas was significantly higher than normal corneas. By the end of the organ culture, the cornea thickness of normal corneas remained at its normal physiological level at 577.20 ± 19.3 µm, while the thickness of the wound corneas was 654.6 ± 31.7 µm. The difference in corneal thickness between the normal group and the wounded group was 112.4 ± 6.9 µm, with a significant difference of p < 0.05, n = 3 biological replicates. Three pairs of corneas were used in this experiment. The donor age and gender were 85 years old female, 79 years old female, and 74 years old male. (B) Representative images (n = 3 biological repeats) of the human cornea of ex vivo organ culture, normal and endothelium wound corneas. Normal corneal showed intact DM and corneal endothelium under a dissecting microscope. (C) The central 8.5 mm DM and endothelium were peeled to create a round corneal endothelium wound. The dashed circle showed the wound area with trypan blue staining. (D) The human cornea remained clear and transparent on day 28, showing a smooth surface with regular light reflection. (E) Endothelium wounded cornea showed an irregular light reflection of the corneal surface with epithelium edema on day 28.
Figure 9.
 
The normal and wounded corneal thickness in an ex vivo anterior chamber infusion model. (A) Three pairs of human corneas were divided into two groups: normal cornea and wounded cornea. Endothelium wounded cornea was made by removing the central 8.5 mm of DM and corneal endothelium. The thickness of wounded corneas was significantly higher than normal corneas. By the end of the organ culture, the cornea thickness of normal corneas remained at its normal physiological level at 577.20 ± 19.3 µm, while the thickness of the wound corneas was 654.6 ± 31.7 µm. The difference in corneal thickness between the normal group and the wounded group was 112.4 ± 6.9 µm, with a significant difference of p < 0.05, n = 3 biological replicates. Three pairs of corneas were used in this experiment. The donor age and gender were 85 years old female, 79 years old female, and 74 years old male. (B) Representative images (n = 3 biological repeats) of the human cornea of ex vivo organ culture, normal and endothelium wound corneas. Normal corneal showed intact DM and corneal endothelium under a dissecting microscope. (C) The central 8.5 mm DM and endothelium were peeled to create a round corneal endothelium wound. The dashed circle showed the wound area with trypan blue staining. (D) The human cornea remained clear and transparent on day 28, showing a smooth surface with regular light reflection. (E) Endothelium wounded cornea showed an irregular light reflection of the corneal surface with epithelium edema on day 28.
Normal Endothelial Cells Remained Alive After 28 Days of Organ Culture, Whereas Cells Were Found in Endothelium Wound Corneas
At the end of the organ culture, corneas were stained with Live/Dead assay to visualize the corneal endothelial cell viability. For normal corneas, most endothelial cells were still alive after 28 days of organ culture, with a round cell nucleus, small cell size and compact cell arrangement. However, for the endothelium-wounded corneas, large numbers of cells showed on the posterior surface of the cornea in the wound area. Their cell bodies were also found to be larger than those of normal endothelial cells, together with disorganized cell arrangement (Fig. 10). 
Figure 10.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture, normal corneas and endothelium wound corneas. Representative confocal microscopy images (n = 3 biological repeats) showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Cell nuclei were counter-stained with Hoechst and emitted blue fluorescence. (AD) Most cells were still alive after 28 days of organ culture for the normal corneas. (EH) Large numbers of live cells were observed in the wounded area, but the cell body was larger than normal endothelial cells, with a disorganized cell arrangement. Scale bar: 50 µm.
Figure 10.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture, normal corneas and endothelium wound corneas. Representative confocal microscopy images (n = 3 biological repeats) showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Cell nuclei were counter-stained with Hoechst and emitted blue fluorescence. (AD) Most cells were still alive after 28 days of organ culture for the normal corneas. (EH) Large numbers of live cells were observed in the wounded area, but the cell body was larger than normal endothelial cells, with a disorganized cell arrangement. Scale bar: 50 µm.
Cells Were Found in the Endothelium Wound Cornea but They Did Not Express Endothelial Cell Markers, ZO-1 and Na/K ATPase
After 28 days of organ culture, both groups of the corneas were taken out of the ex vivo model for ZO-1, Na/K ATPase, and F-actin staining. Normal corneas showed enlarged and polygonal corneal endothelial cells expressing ZO-1 on the cell-cell borders. To create an endothelium wound model to reflect endothelial decompensation, a central 8.5 mm DM was manually stripped. Immunostaining showed that there was no ZO-1 expression on the cell-cell border. On the contrary, large numbers of cells were observed on the posterior surface of the cornea in the wounded area. These cells did not show typical endothelial cell markers, instead showing the irregular shape of the cell nucleus (Figs. 11A–F). 
Figure 11.
 
Immunofluorescence staining of flat-mounted normal and endothelium wound cornea after 28 days of organ culture. Representative confocal microscopy images (n = 3 biological repeats) showed the results of ZO-1, Na/K ATPase and F-actin staining. (AC) Normal corneas showed enlarged and polygonal corneal endothelial cells expressing ZO-1 on the cell-cell borders. (DF) Large numbers of cells were found on the wound area for the endothelium-wounded cornea, with no ZO-1 expression. (GI) Cells expressed Na/K ATPase on the cell-cell borders and the cell membrane in normal corneas. (JL) There was no Na/K ATPase expression on the endothelium wound corneas. (MO) For the normal cornea, the F-actin cytoskeleton showed the typical polygonal shape of the endothelial cell. However, (PR) a disorganized f-actin cytoskeleton was on the wound area with stress fibers across the cell bodies. Nuclei were counterstained with Hoechst 33342. Scale bar: 50 µm.
Figure 11.
 
Immunofluorescence staining of flat-mounted normal and endothelium wound cornea after 28 days of organ culture. Representative confocal microscopy images (n = 3 biological repeats) showed the results of ZO-1, Na/K ATPase and F-actin staining. (AC) Normal corneas showed enlarged and polygonal corneal endothelial cells expressing ZO-1 on the cell-cell borders. (DF) Large numbers of cells were found on the wound area for the endothelium-wounded cornea, with no ZO-1 expression. (GI) Cells expressed Na/K ATPase on the cell-cell borders and the cell membrane in normal corneas. (JL) There was no Na/K ATPase expression on the endothelium wound corneas. (MO) For the normal cornea, the F-actin cytoskeleton showed the typical polygonal shape of the endothelial cell. However, (PR) a disorganized f-actin cytoskeleton was on the wound area with stress fibers across the cell bodies. Nuclei were counterstained with Hoechst 33342. Scale bar: 50 µm.
Similar results were also observed for the Na/K ATPase staining. Large numbers of cells were found on the endothelium wounded area, but none of them showed Na/K ATPase expression on the cell-cell borders. F-actin staining showed that the cytoskeleton of the cells was randomly distributed throughout the cytoplasm with disorganized stress fibers across the cell bodies in the wound area (Figs. 11G–L). On the other hand, F-actin staining results showed that the normal cornea showed an organized cytoskeleton at the cell borders, and the cells retained the characteristic polygonal shape of the endothelial cell (Figs. 11M–R). 
Endothelium Wound Cornea Lost Central DM and Endothelial Cells
Cross-sectional H&E staining showed that intact DM and endothelial cells were preserved in the normal cornea. The central 8.5 mm DM was peeled to create an endothelium wound area, and the H&E staining confirmed this wound to see the end point of peeled DM on the peripheral cornea. Also, multilayers of epithelial cells were present on the anterior surface of the cornea after 28 days of ex vivo organ culture. The thickness of epithelium in the wound cornea was much thinner than in the normal cornea (Figs. 12A–D). 
Figure 12.
 
Cross-sectional H&E staining and immunostaining of collagen IV in normal cornea and endothelium wound cornea after 28 days of organ culture. (A) Intact DM was on the posterior surface of the cornea. (B) Multilayers of epithelial cells were observed on the anterior surface of the cornea. (C) An endothelium wound was created by peeling the central 8.5 mm DM. (D) The corneal epithelium in the wound cornea was thinner than that of the normal cornea. Scale bar: 100 µm. (EG) Normal cornea showed intact collagen IV expressions on DM across the peripheral to central cornea, with a single cell layer of endothelial cells. (IK) A central 8.5 mm cornea defect in DM was confirmed by labelling collagen IV of DM. Cells were found on the wound area, residing in the posterior stroma. (H, L) Images showed collagen IV expression on the epithelium's basement membrane in normal and wound corneas. Also, the thickness of the epithelium in the wound cornea was thinner than in the normal cornea. Representative images, n = 3 biological replicates. Cell nuclei were counterstained with Hoechst. Scale bar: 50 µm.
Figure 12.
 
Cross-sectional H&E staining and immunostaining of collagen IV in normal cornea and endothelium wound cornea after 28 days of organ culture. (A) Intact DM was on the posterior surface of the cornea. (B) Multilayers of epithelial cells were observed on the anterior surface of the cornea. (C) An endothelium wound was created by peeling the central 8.5 mm DM. (D) The corneal epithelium in the wound cornea was thinner than that of the normal cornea. Scale bar: 100 µm. (EG) Normal cornea showed intact collagen IV expressions on DM across the peripheral to central cornea, with a single cell layer of endothelial cells. (IK) A central 8.5 mm cornea defect in DM was confirmed by labelling collagen IV of DM. Cells were found on the wound area, residing in the posterior stroma. (H, L) Images showed collagen IV expression on the epithelium's basement membrane in normal and wound corneas. Also, the thickness of the epithelium in the wound cornea was thinner than in the normal cornea. Representative images, n = 3 biological replicates. Cell nuclei were counterstained with Hoechst. Scale bar: 50 µm.
Labelling collagen IV (COL4) was used to confirm the DM distribution of the cornea after organ culture. For the normal cornea, intact DM with COL4 expression was observed on the posterior surface of the cornea, from the periphery to the central cornea. A monolayer of cornea endothelial cells was attached to the DM in the normal cornea. 
On the contrary, COL4 expression was missing on the posterior surface of the wound cornea, from the paracentral to the central cornea. This confirmed that a cornea defect in DM was created after peeling the central 8.5 mm DM. Large numbers of cells were found in the posterior stroma in the wound area. For the epithelium, a multilayer of epithelial cells was observed on the anterior surface of the cornea in both groups, with intact COL4 expression on the basal side of the epithelium. However, the thickness of epithelium in the wound cornea was thinner than the thickness of epithelium in the normal cornea (Figs. 12E–L). 
Discussion
The major function of corneal endothelial cells is to maintain cornea transparency by regulating the hydration level of the corneal stroma. Previous research has shown that corneal endothelial cells can be expanded and cultured in vitro to propagate large numbers of cells for tissue engineering. Cultured corneal endothelial cells can retain typical cell morphology with polygonal cell shape and compact cell-cell contacts and have the functional gene and protein expressions, such as tight junction ZO-1 and transmembrane protein Na/K ATPase.17 However, whether in vitro cultured corneal endothelial cells can function properly in vivo to maintain normal corneal thickness and the feasibility of transplanting a tissue-engineered corneal endothelium in vivo still need to be fully investigated. 
Many animal models, such as rabbits, rats, primates, and felines, have been involved in corneal endothelial research.22 However, these models are limited. For example, the corneal endothelial cells of rabbits and rats have been observed to possess proliferative ability in vivo.23 Although the proliferative capacity of corneal endothelial cells of feline and primates is similar to humans with limited cell dividing ability in vivo,24,25 the research cost and the maintenance of large animal research could be very expensive and require special regulation. Alternatively, developing an organ culture system has attempted to replicate better the native human tissue microenvironment for cell therapy studies.2628 Rolev et al.28 developed an air interface corneal organ culture model by mounting a human cornea onto an artificial anterior chamber. However, the results showed that their model could sustain cornea tissue with the normal physiology of cornea thickness for only 12 hours. 
This study successfully established a new ex vivo anterior chamber infusion model, with continuous infusion of culture medium to support corneal endothelial cells and corneal epithelial cells simultaneously and over a prolonged period. We modified previous organ culture models in several ways to mimic the in vivo environment better. First, instead of manually injecting the culture medium into the anterior chamber, the chamber was connected to a culture medium pumping system, including a tubing circuit, a peristaltic pump, and a culture medium reservoir. The pumping system creates a continuous flow of culture medium, which provides a 24-hour infusion to the corneal endothelial cells and takes away the metabolic wastes back to the reservoir. Second, a bespoke culture medium container ring was attached on the epithelial side of the anterior chamber. The container ring retained the epithelial culture medium on the cornea's anterior surface to culture and maintain epithelial cells. Together with endothelial cell culture inside the chamber, this new system fulfilled the co-culture model for both epithelial and endothelial cells. On the one hand, the epithelial culture medium keeps the ocular surface moisture to prevent overwater evaporation from the stroma. Furthermore, it maintains the corneal limbus to keep the integrity of the epithelium barrier function. Third, the reservoir was placed 200 mm above the chamber to create a hydrostatic pressure to maintain the anterior chamber's pressure and the anterior segment structure of the eye. 
Corneas showed edema at the beginning of the experiment, but the cornea swelling subsided after being cultured in our ex vivo model. The gradual decrease in cornea thickness at the beginning of the organ culture indicated that the corneal endothelial cells began to resume their barrier function to pump water out of the stroma and to reduce the corneal thickness. The corneal thickness returned to normal within one day for the human cornea with intact endothelium. Our model maintained the cornea thickness at a normal physiological level for 28 days without showing edema, which is longer than previous published ex vivo cornea organ culture time.28 Camera images also indicated that the human cornea was still clear with only a few punctate defects on the epithelium at the end of the experiment. Together, these results suggested that corneal endothelial cells can resume their function in our organ culture model, and that their function can be indirectly evaluated by measuring the corneal thickness changes. In addition, an endothelial wound model was also created by manually peeling the DM. This caused persistent corneal edema with an increase in corneal thickness rising to more than 650 µm. Without corneal endothelial cells, the cornea displays persistent swelling and increased thickness. These findings further validated the relationship between endothelial cell function and corneal thickness changes supported by the “normal” and “wounded” models. 
The corneal limbus was included in our ex vivo model to maintain corneal epithelium. The corneal limbus is the transition zone between the cornea and the sclera.29,30 The limbal epithelial cells are widely believed to harbor groups of highly proliferative stem cells, which can regenerate themselves to replenish cell loss.30 The proliferative function of the limbus might have helped support epithelial cell regeneration and maintenance of the cornea overall during long-term organ culture in our model. Our histological results could also support this argument where multilayers of epithelial cells were observed on the surface of the cornea after 28 days of organ culture. Sample preparation for the cryo-sectional staining might account for the epithelium thinning observed in the endothelium wound cornea. Epithelium edema was observed in the endothelium wound corneas during the organ culture. The unstable multilayers of epithelial cells might be partially missing during sample embedding and sectioning. 
At the end of the experiment, human corneas were taken out of the organ culture chamber for cell viability assessment and immunofluorescence staining of typical endothelial cell markers. After 28 days of organ culture, most cells remained alive and expressed ZO- and Na/K ATPase. Tight junction protein ZO-1 and transmembrane protein Na/K ATPase are the major components of corneal endothelial cells used to form a fluid barrier controlling the hydration level of the cornea. Our model successfully maintained corneal endothelial cell viability and characteristic cell marker expression. These attributes may have contributed to the maintenance of the cells’ barrier function, thereby sustaining the normal physiological thickness of the cornea in an ex vivo environment. 
After 28 days of organ culture, endothelial cells were found to be missing, perhaps indicating the limits of the model in supporting endothelial cells. The cell monolayer defects showed a moth-eaten appearance, with dead cells at the edge of the defect. Endothelial cell damage might derive from the donor tissue preparation itself, such as the harvesting of the cornea, the storage of the tissue in culture media, or tissue folding during the preparation.3133 Experimental handling is another reason that causes cell damage and defects. The dead cells showing on the edge of the defects might also come from cell apoptosis or cell ageing during organ culture. Previous research has shown that apoptotic cells were identified in the corneal endothelium of human organ-cultured corneas.31 This might account for the endothelial cell loss in our long-term ex vivo infusion model. In addition, our results demonstrated that despite small areas of endothelium defect, the remaining endothelial cells could still function properly to keep the cornea clear and sustain normal corneal thickness. Apart from the small area of cell defects, it was also observed that the cell morphology changes at the end of the organ culture period. Specifically, ZO-1 and Na/K ATPase staining showed that cell area increased and cell hexagonality was lost. The cells displayed a polygonal cell shape and irregular cell arrangement. The variation in cell size and cell shape of corneal endothelial cells in our models are similar to the in vivo morphology changes to corneal endothelial cells with cell migration and expansion to compensate for cell loss.34 This suggests that the corneal endothelial cells of the cadaver cornea might have undergone cell migration and expansion to compensate for small cell defects in our ex vivo model. 
It has been widely believed that human corneal endothelial cells have only limited regenerative ability. The mechanism of why corneal endothelial cells cannot divide in vivo is still unclear. Possible reasons include cell-cell contact-dependent inhibition, lack of effective growth factor stimulation and TGF-β2 suppression.35 Previous studies observed that human corneal endothelial cells still have the regenerative ability to cover endothelium wounds caused by severe ocular burns. The wound area underwent re-endothelialization through cell proliferation and migration from the far periphery endothelial cells of the posterior limbus.36 In our wound model, the central DM and endothelium were manually peeled to create an 8.5 mm round endothelium and DM defect, while the periphery DM and endothelial cells were preserved. Based on the immunostaining results, there was no sign that the periphery endothelial cells can migrate or proliferate during wound healing. A possible explanation could be that the DM was removed. DM has been considered to have an influence on corneal endothelial cell regeneration.37,38 Previous experimental research compared the influence of DM on corneal endothelial cell regeneration in the rabbit.39 Their results showed that the adjacent corneal endothelial cells underwent cell transition and cell migration to cover the cell defect with intact DM. In contrast, the cell migration was limited in the cell defect with DM removal. This indicated that DM plays an important role for corneal endothelial cell regeneration by providing a scaffold for corneal endothelial migration and proliferation.39 
In addition, removal of the DM will also expose the underlying posterior stroma to the aqueous humor. The keratocytes in the corneal stroma can be activated and transformed into myofibroblasts, if in contact with growth factors in the aqueous humour.4042 Previous research has shown that cells were found on the posterior surface of the stroma in rabbit corneal endothelium wound model after 14 days of DM stripping. The cells were not originated from corneal endothelial cells and did not express ZO-1 and Na/K ATPase markers.39 Similar results were also observed in our ex vivo endothelium wound experiment, where cells were found on the posterior surface of the cornea in the wound area. These cells did not show endothelial cell markers, ZO-1 and Na/K ATPase but instead showed fibroblast cell morphology. A possible explanation is that these cells were differentiated from the keratocytes in the posterior stroma. The growth factors in the endothelium culture medium trigger the posterior stromal cell to differentiate into myofibroblasts in our ex vivo organ culture model. 
The results of this study showed that the ex vivo anterior chamber infusion model could sustain human cornea tissues in an organ culture system for one month. The endothelium wound corneas showed persistent corneal edema and increased corneal thickness, 654.6 ± 31.7 µm, on 28 days of organ culture. The corneal thickness of the wound model is close to the thickness changes in Fuchs’ endothelial dystrophy, with a reported central corneal thickness of 614 ± 57 µm.43 This indicates that this wound model could reflect the corneal thickness changes in endothelial dysfunction patients. Thus this model can be useful for studying endothelium wound healing experiments and an alternative to the animal model for assessing cell therapy transplantation. 
Acknowledgments
The authors thank the National Health Service Blood and Transplant (NHSBT) Tissue and Eye services for support in obtaining consent and providing the tissue (corneas) from UK donors. We also thank Yann Bouremel and Sir Peng Khaw at UCL Institute of Ophthalmology for the 3D printing work. The publication costs are supported only by the ARVO's Publication Financial Assistance Program. 
Disclosure: M.-C. Tsai, None; A. Kureshi, None; J.T. Daniels, None 
References
Price MO, Mehta JS, Jurkunas UV, Price FW, Jr. Corneal endothelial dysfunction: Evolving understanding and treatment options. Prog Retin Eye Res. 2021; 82: 100904. [CrossRef] [PubMed]
Peh GSL, Ong HS, Adnan K, et al. Functional Evaluation of Two Corneal Endothelial Cell-Based Therapies: Tissue-Engineered Construct and Cell Injection. Sci Rep. 2019; 9: 6087. [CrossRef] [PubMed]
Sun P, Shen L, Zhang C, Du L, Wu X. Promoting the expansion and function of human corneal endothelial cells with an orbital adipose-derived stem cell-conditioned medium. Stem Cell Res Ther. 2017; 8: 287. [CrossRef] [PubMed]
Meekins LC, Rosado-Adames N, Maddala R, Zhao JJ, Rao PV, Afshari NA. Corneal Endothelial Cell Migration and Proliferation Enhanced by Rho Kinase (ROCK) Inhibitors in In Vitro and In Vivo Models. Invest Ophthalmol Vis Sci. 2016; 57: 6731–6738. [CrossRef] [PubMed]
Bi YL, Zhou Q, Du F, Wu MF, Xu GT, Sui GQ. Regulation of functional corneal endothelial cells isolated from sphere colonies by Rho-associated protein kinase inhibitor. Exp Ther Med. 2013; 5: 433–437. [CrossRef] [PubMed]
Hatou S, Higa K, Inagaki E, et al. Validation of Na,K-ATPase pump function of corneal endothelial cells for corneal regenerative medicine. Tissue Eng Part C Methods. 2013; 19: 901–910. [CrossRef] [PubMed]
Hatou S, Yamada M, Akune Y, et al. Role of Insulin in Regulation of Na+-/K+-Dependent ATPase Activity and Pump Function in Corneal Endothelial Cells. Invest Ophthalmol Vis Sci. 2010; 51: 3935–3942. [CrossRef] [PubMed]
Srinivasan B, Kolli AR, Esch MB, Abaci HE, Shuler ML, Hickman JJ. TEER Measurement Techniques for In Vitro Barrier Model Systems. Journal of Laboratory Automation. 2015; 20: 107–126. [CrossRef] [PubMed]
Mimura T, Yamagami S, Yokoo S, et al. Cultured human corneal endothelial cell transplantation with a collagen sheet in a rabbit model. Invest Ophthalmol Vis Sci. 2004; 45: 2992–2997. [CrossRef] [PubMed]
Tretola M, Bee G, Dohme-Meier F, Silacci P. Review: Harmonised in vitro digestion and the Ussing chamber for investigating the effects of polyphenols on intestinal physiology in monogastrics and ruminants. animal. 2023; 17: 100785. [CrossRef] [PubMed]
Pei W, Chen J, Wu W, Wei W, Yu Y, Feng Y. Comparison of the rabbit and human corneal endothelial proteomes regarding proliferative capacity. Exp Eye Res. 2021; 209: 108629. [CrossRef] [PubMed]
Loiseau A, Raîche-Marcoux G, Maranda C, Bertrand N, Boisselier E. Animal Models in Eye Research: Focus on Corneal Pathologies. Int J Mol Sci. 2023; 24.
Sanchez I, Martin R, Ussa F, Fernandez-Bueno I. The parameters of the porcine eyeball. Graefes Arch Clin Exp Ophthalmol. 2011; 249: 475–482. [CrossRef] [PubMed]
Riau AK, Tan NYS, Angunawela RI, Htoon HM, Chaurasia SS, Mehta JS. Reproducibility and age-related changes of ocular parametric measurements in rabbits. BMC Vet Res. 2012; 8: 138. [CrossRef] [PubMed]
Bozkir G, Bozkir M, Dogan H, Aycan K, Güler B. Measurements of axial length and radius of corneal curvature in the rabbit eye. Acta Med Okayama. 1997; 51: 9–11. [PubMed]
Parikumar P, Haraguchi K, Ohbayashi A, Senthilkumar R, Abraham SJ. Successful transplantation of in vitro expanded human cadaver corneal endothelial precursor cells on to a cadaver bovine's eye using a nanocomposite gel sheet. Curr Eye Res. 2014; 39: 522–526. [CrossRef] [PubMed]
Levis HJ, Peh GS, Toh KP, et al. Plastic compressed collagen as a novel carrier for expanded human corneal endothelial cells for transplantation. PLoS One. 2012; 7: e50993. [CrossRef] [PubMed]
Yoeruek E, Bayyoud T, Maurus C, et al. Decellularization of porcine corneas and repopulation with human corneal cells for tissue-engineered xenografts. Acta Ophthalmol. 2012; 90: e125–131. [PubMed]
Bayyoud T, Thaler S, Hofmann J, et al. Decellularized bovine corneal posterior lamellae as carrier matrix for cultivated human corneal endothelial cells. Curr Eye Res. 2012; 37: 179–186. [CrossRef] [PubMed]
Amano S, Mimura T, Yamagami S, Osakabe Y, Miyata K. Properties of corneas reconstructed with cultured human corneal endothelial cells and human corneal stroma. Jpn J Ophthalmol. 2005; 49: 448–452. [CrossRef] [PubMed]
Choi JS, Williams JK, Greven M, et al. Bioengineering endothelialized neo-corneas using donor-derived corneal endothelial cells and decellularized corneal stroma. Biomaterials. 2010; 31: 6738–6745. [CrossRef] [PubMed]
Rolev K, Coussons P, King L, Rajan M. Experimental models of corneal endothelial cell therapy and translational challenges to clinical practice. Exp Eye Res. 2019; 188: 107794. [CrossRef] [PubMed]
Van Horn DL, Sendele DD, Seideman S, Buco PJ. Regenerative capacity of the corneal endothelium in rabbit and cat. Invest Ophthalmol Vis Sci. 1977; 16: 597–613. [PubMed]
Van Horn DL, Hyndiuk RA. Endothelial wound repair in primate cornea. Exp Eye Res. 1975; 21: 113–124. [CrossRef] [PubMed]
Huang PT, Nelson LR, Bourne WM. The morphology and function of healing cat corneal endothelium. Invest Ophthalmol Vis Sci. 1989; 30: 1794–1801. [PubMed]
Zhao B, Cooper LJ, Brahma A, MacNeil S, Rimmer S, Fullwood NJ. Development of a three-dimensional organ culture model for corneal wound healing and corneal transplantation. Invest Ophthalmol Vis Sci. 2006; 47: 2840–2846. [CrossRef] [PubMed]
Patel SV, Bachman LA, Hann CR, Bahler CK, Fautsch MP. Human corneal endothelial cell transplantation in a human ex vivo model. Invest Ophthalmol Vis Sci. 2009; 50: 2123–2131. [CrossRef] [PubMed]
Rolev K, OʼDonovan DG, Coussons P, King L, Rajan MS. Feasibility Study of Human Corneal Endothelial Cell Transplantation Using an In Vitro Human Corneal Model. Cornea. 2018; 37: 778–784. [CrossRef] [PubMed]
Van Buskirk EM. The anatomy of the limbus. Eye (Lond). 1989; 3(Pt. 2): 101–108. [PubMed]
Tseng SCG. Concept and application of limbal stem cells. Eye. 1989; 3: 141–157. [CrossRef] [PubMed]
Albon J, Tullo AB, Aktar S, Boulton ME. Apoptosis in the endothelium of human corneas for transplantation. Invest Ophthalmol Vis Sci. 2000; 41: 2887–2893. [PubMed]
Bhogal M, Balda MS, Matter K, Allan BD. Global cell-by-cell evaluation of endothelial viability after two methods of graft preparation in Descemet membrane endothelial keratoplasty. Br J Ophthalmol. 2016; 100: 572–578. [CrossRef] [PubMed]
Bhogal M, Lwin CN, Seah XY, et al. Real-time assessment of corneal endothelial cell damage following graft preparation and donor insertion for DMEK. PLoS One. 2017; 12: e0184824. [CrossRef] [PubMed]
Bourne WM. Biology of the corneal endothelium in health and disease. Eye (Lond). 2003; 17: 912–918. [CrossRef] [PubMed]
Joyce NC. Proliferative capacity of corneal endothelial cells. Exp Eye Res. 2012; 95: 16–23. [CrossRef] [PubMed]
Choi SO, Jeon HS, Hyon JY, et al. Recovery of corneal endothelial cells from periphery after injury. PLoS One. 2015; 10: e0138076. [CrossRef] [PubMed]
Garcerant D, Hirnschall N, Toalster N, Zhu M, Wen L, Moloney G. Descemet's stripping without endothelial keratoplasty. Curr Opin Ophthalmol. 2019; 30: 275–285. [CrossRef] [PubMed]
Park S, Leonard BC, Raghunathan VK, et al. Animal models of corneal endothelial dysfunction to facilitate development of novel therapies. Ann Transl Med. 2020; 9: 1271. [CrossRef]
Chen J, Li Z, Zhang L, et al. Descemet's membrane supports corneal endothelial cell regeneration in rabbits. Sci Rep. 2017; 7: 6983. [CrossRef] [PubMed]
Medeiros CS, Saikia P, de Oliveira RC, Lassance L, Santhiago MR, Wilson SE. Descemet's membrane modulation of posterior corneal fibrosis. Invest Ophthalmol Vis Sci. 2019; 60: 1010–1020. [CrossRef] [PubMed]
Medeiros CS, Marino GK, Santhiago MR, Wilson SE. The corneal basement membranes and stromal fibrosis. Invest Ophthalmol Vis Sci. 2018; 59: 4044–4053. [CrossRef] [PubMed]
Wilson SE. Corneal myofibroblast biology and pathobiology: Generation, persistence, and transparency. Exp Eye Res. 2012; 99: 78–88. [CrossRef] [PubMed]
Repp DJ, Hodge DO, Baratz KH, McLaren JW, Patel SV. Fuchs' endothelial corneal dystrophy: subjective grading versus objective grading based on the central-to-peripheral thickness ratio. Ophthalmology. 2013; 120: 687–694. [CrossRef] [PubMed]
Figure 1.
 
The diagram illustrates the experimental design of establishing a human cornea ex vivo anterior chamber infusion model. Three individual human corneas were used to establish a human cornea ex vivo anterior chamber infusion model. N = 3 biological replicates. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay and flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase.
Figure 1.
 
The diagram illustrates the experimental design of establishing a human cornea ex vivo anterior chamber infusion model. Three individual human corneas were used to establish a human cornea ex vivo anterior chamber infusion model. N = 3 biological replicates. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay and flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase.
Figure 2.
 
Schematic representation of the ex vivo anterior chamber infusion model. This infusion model comprises a reservoir bottle, a peristaltic pump, and an artificial anterior chamber. The anterior chamber infusion system uses a peristaltic pump to create a flow so the culture medium can flow back to the culture medium reservoir bottle. The arrowhead indicates the direction of the culture medium flow. Cornea tissue is mounted on the anterior chamber and locked with a tissue retainer to create a closed chamber for culture medium infusion. The inlet and outlet of the ports were connected to a peristaltic pump (100 µL/min) and a culture medium reservoir bottle with connecting tubing to complete the flow circuit. A 0.2 µm filter was attached to the reservoir bottle to enable air exchange.
Figure 2.
 
Schematic representation of the ex vivo anterior chamber infusion model. This infusion model comprises a reservoir bottle, a peristaltic pump, and an artificial anterior chamber. The anterior chamber infusion system uses a peristaltic pump to create a flow so the culture medium can flow back to the culture medium reservoir bottle. The arrowhead indicates the direction of the culture medium flow. Cornea tissue is mounted on the anterior chamber and locked with a tissue retainer to create a closed chamber for culture medium infusion. The inlet and outlet of the ports were connected to a peristaltic pump (100 µL/min) and a culture medium reservoir bottle with connecting tubing to complete the flow circuit. A 0.2 µm filter was attached to the reservoir bottle to enable air exchange.
Figure 3.
 
Pictures of the ex vivo anterior chamber infusion model. (A) The Baron artificial anterior chamber with 3D-printed medium container ring. (B) The ex vivo anterior chamber infusion model was connected to a peristaltic pump and placed on a 3D orbital rotating shake in the incubator at 37°C, with 5% CO2 in the air.
Figure 3.
 
Pictures of the ex vivo anterior chamber infusion model. (A) The Baron artificial anterior chamber with 3D-printed medium container ring. (B) The ex vivo anterior chamber infusion model was connected to a peristaltic pump and placed on a 3D orbital rotating shake in the incubator at 37°C, with 5% CO2 in the air.
Figure 4.
 
The diagram illustrates the experimental design of establishing an ex vivo corneal endothelium wound model. Three pairs of human corneas (n = 3 biological replicates) were divided into two groups: normal corneas and endothelium wound corneas. The central 8.5 mm DM and endothelium were peeled to create an endothelium wound. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay, flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase, and cross-sectional staining.
Figure 4.
 
The diagram illustrates the experimental design of establishing an ex vivo corneal endothelium wound model. Three pairs of human corneas (n = 3 biological replicates) were divided into two groups: normal corneas and endothelium wound corneas. The central 8.5 mm DM and endothelium were peeled to create an endothelium wound. Corneal thickness changes were used to assess the corneal endothelial cell function. After 28 days of organ culture, the corneas were sent for live/dead assay, flat-mounted immunofluorescence staining of ZO-1 and Na/K ATPase, and cross-sectional staining.
Figure 5.
 
The human corneal thickness of ex vivo anterior chamber infusion model. Human corneas with intact corneal endothelium, normal group, were cultured and maintained in our models for 28 days. The initial thickness was 934.73 ± 52.76 µm, and the number decreased to the normal physiological level of 571.20 ± 38.05 µm within one day. Then, the corneas maintained a thickness of 570.53 ± 23.28 µm until day 28. Cornea thickness was measured by ultrasonic handheld pachymeter, with five measurements for each cornea. N = 3 biological replicates. Donor age and gender were 72 years old male, 74 years old female, and 75 years old male.
Figure 5.
 
The human corneal thickness of ex vivo anterior chamber infusion model. Human corneas with intact corneal endothelium, normal group, were cultured and maintained in our models for 28 days. The initial thickness was 934.73 ± 52.76 µm, and the number decreased to the normal physiological level of 571.20 ± 38.05 µm within one day. Then, the corneas maintained a thickness of 570.53 ± 23.28 µm until day 28. Cornea thickness was measured by ultrasonic handheld pachymeter, with five measurements for each cornea. N = 3 biological replicates. Donor age and gender were 72 years old male, 74 years old female, and 75 years old male.
Figure 6.
 
Image of the human cornea after 28 days of ex vivo organ culture. (A) The human cornea remained clear and transparent on day 28. (B) Small punctate epithelial defects were observed after staining the cornea with 2% fluorescein.
Figure 6.
 
Image of the human cornea after 28 days of ex vivo organ culture. (A) The human cornea remained clear and transparent on day 28. (B) Small punctate epithelial defects were observed after staining the cornea with 2% fluorescein.
Figure 7.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture. Confocal microscopy (representative of n = 3 biological replicates) images showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Most of the cells were still alive after 28 days of organ culture. However, some endothelial cells were missing, with small numbers of dead cells on the edge of the defect (arrowhead). Some large defects were also observed, displaying a moth-eaten appearance. Scale bar: 50 µm.
Figure 7.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture. Confocal microscopy (representative of n = 3 biological replicates) images showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Most of the cells were still alive after 28 days of organ culture. However, some endothelial cells were missing, with small numbers of dead cells on the edge of the defect (arrowhead). Some large defects were also observed, displaying a moth-eaten appearance. Scale bar: 50 µm.
Figure 8.
 
Immunofluorescence staining of ZO-1 and Na/K ATPase on the flat-mounted cornea. (A, C) ZO-1 was labelled with green fluorescence and displayed a clear intercellular outline on the cell-cell borders. Some cells became enlarged, with loss of cell hexagonality, and some cells were missing, leaving behind a small area of defect. (B, D) Na/K ATPase was labelled with green fluorescence and displayed a clear outline of cell membranes. Like ZO-1 staining, some endothelial cells were lost, leaving a small area of cell defect behind. Nuclei were counterstained with Hoechst 33342. Representative images, n = 3 biological replicates. Scale bar: 50 µm.
Figure 8.
 
Immunofluorescence staining of ZO-1 and Na/K ATPase on the flat-mounted cornea. (A, C) ZO-1 was labelled with green fluorescence and displayed a clear intercellular outline on the cell-cell borders. Some cells became enlarged, with loss of cell hexagonality, and some cells were missing, leaving behind a small area of defect. (B, D) Na/K ATPase was labelled with green fluorescence and displayed a clear outline of cell membranes. Like ZO-1 staining, some endothelial cells were lost, leaving a small area of cell defect behind. Nuclei were counterstained with Hoechst 33342. Representative images, n = 3 biological replicates. Scale bar: 50 µm.
Figure 9.
 
The normal and wounded corneal thickness in an ex vivo anterior chamber infusion model. (A) Three pairs of human corneas were divided into two groups: normal cornea and wounded cornea. Endothelium wounded cornea was made by removing the central 8.5 mm of DM and corneal endothelium. The thickness of wounded corneas was significantly higher than normal corneas. By the end of the organ culture, the cornea thickness of normal corneas remained at its normal physiological level at 577.20 ± 19.3 µm, while the thickness of the wound corneas was 654.6 ± 31.7 µm. The difference in corneal thickness between the normal group and the wounded group was 112.4 ± 6.9 µm, with a significant difference of p < 0.05, n = 3 biological replicates. Three pairs of corneas were used in this experiment. The donor age and gender were 85 years old female, 79 years old female, and 74 years old male. (B) Representative images (n = 3 biological repeats) of the human cornea of ex vivo organ culture, normal and endothelium wound corneas. Normal corneal showed intact DM and corneal endothelium under a dissecting microscope. (C) The central 8.5 mm DM and endothelium were peeled to create a round corneal endothelium wound. The dashed circle showed the wound area with trypan blue staining. (D) The human cornea remained clear and transparent on day 28, showing a smooth surface with regular light reflection. (E) Endothelium wounded cornea showed an irregular light reflection of the corneal surface with epithelium edema on day 28.
Figure 9.
 
The normal and wounded corneal thickness in an ex vivo anterior chamber infusion model. (A) Three pairs of human corneas were divided into two groups: normal cornea and wounded cornea. Endothelium wounded cornea was made by removing the central 8.5 mm of DM and corneal endothelium. The thickness of wounded corneas was significantly higher than normal corneas. By the end of the organ culture, the cornea thickness of normal corneas remained at its normal physiological level at 577.20 ± 19.3 µm, while the thickness of the wound corneas was 654.6 ± 31.7 µm. The difference in corneal thickness between the normal group and the wounded group was 112.4 ± 6.9 µm, with a significant difference of p < 0.05, n = 3 biological replicates. Three pairs of corneas were used in this experiment. The donor age and gender were 85 years old female, 79 years old female, and 74 years old male. (B) Representative images (n = 3 biological repeats) of the human cornea of ex vivo organ culture, normal and endothelium wound corneas. Normal corneal showed intact DM and corneal endothelium under a dissecting microscope. (C) The central 8.5 mm DM and endothelium were peeled to create a round corneal endothelium wound. The dashed circle showed the wound area with trypan blue staining. (D) The human cornea remained clear and transparent on day 28, showing a smooth surface with regular light reflection. (E) Endothelium wounded cornea showed an irregular light reflection of the corneal surface with epithelium edema on day 28.
Figure 10.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture, normal corneas and endothelium wound corneas. Representative confocal microscopy images (n = 3 biological repeats) showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Cell nuclei were counter-stained with Hoechst and emitted blue fluorescence. (AD) Most cells were still alive after 28 days of organ culture for the normal corneas. (EH) Large numbers of live cells were observed in the wounded area, but the cell body was larger than normal endothelial cells, with a disorganized cell arrangement. Scale bar: 50 µm.
Figure 10.
 
Live/Dead cell viability assay of human cornea endothelial cells after 28 days of ex vivo organ culture, normal corneas and endothelium wound corneas. Representative confocal microscopy images (n = 3 biological repeats) showed that the dead cells emitted red fluorescence, and the live cells emitted green fluorescence. Cell nuclei were counter-stained with Hoechst and emitted blue fluorescence. (AD) Most cells were still alive after 28 days of organ culture for the normal corneas. (EH) Large numbers of live cells were observed in the wounded area, but the cell body was larger than normal endothelial cells, with a disorganized cell arrangement. Scale bar: 50 µm.
Figure 11.
 
Immunofluorescence staining of flat-mounted normal and endothelium wound cornea after 28 days of organ culture. Representative confocal microscopy images (n = 3 biological repeats) showed the results of ZO-1, Na/K ATPase and F-actin staining. (AC) Normal corneas showed enlarged and polygonal corneal endothelial cells expressing ZO-1 on the cell-cell borders. (DF) Large numbers of cells were found on the wound area for the endothelium-wounded cornea, with no ZO-1 expression. (GI) Cells expressed Na/K ATPase on the cell-cell borders and the cell membrane in normal corneas. (JL) There was no Na/K ATPase expression on the endothelium wound corneas. (MO) For the normal cornea, the F-actin cytoskeleton showed the typical polygonal shape of the endothelial cell. However, (PR) a disorganized f-actin cytoskeleton was on the wound area with stress fibers across the cell bodies. Nuclei were counterstained with Hoechst 33342. Scale bar: 50 µm.
Figure 11.
 
Immunofluorescence staining of flat-mounted normal and endothelium wound cornea after 28 days of organ culture. Representative confocal microscopy images (n = 3 biological repeats) showed the results of ZO-1, Na/K ATPase and F-actin staining. (AC) Normal corneas showed enlarged and polygonal corneal endothelial cells expressing ZO-1 on the cell-cell borders. (DF) Large numbers of cells were found on the wound area for the endothelium-wounded cornea, with no ZO-1 expression. (GI) Cells expressed Na/K ATPase on the cell-cell borders and the cell membrane in normal corneas. (JL) There was no Na/K ATPase expression on the endothelium wound corneas. (MO) For the normal cornea, the F-actin cytoskeleton showed the typical polygonal shape of the endothelial cell. However, (PR) a disorganized f-actin cytoskeleton was on the wound area with stress fibers across the cell bodies. Nuclei were counterstained with Hoechst 33342. Scale bar: 50 µm.
Figure 12.
 
Cross-sectional H&E staining and immunostaining of collagen IV in normal cornea and endothelium wound cornea after 28 days of organ culture. (A) Intact DM was on the posterior surface of the cornea. (B) Multilayers of epithelial cells were observed on the anterior surface of the cornea. (C) An endothelium wound was created by peeling the central 8.5 mm DM. (D) The corneal epithelium in the wound cornea was thinner than that of the normal cornea. Scale bar: 100 µm. (EG) Normal cornea showed intact collagen IV expressions on DM across the peripheral to central cornea, with a single cell layer of endothelial cells. (IK) A central 8.5 mm cornea defect in DM was confirmed by labelling collagen IV of DM. Cells were found on the wound area, residing in the posterior stroma. (H, L) Images showed collagen IV expression on the epithelium's basement membrane in normal and wound corneas. Also, the thickness of the epithelium in the wound cornea was thinner than in the normal cornea. Representative images, n = 3 biological replicates. Cell nuclei were counterstained with Hoechst. Scale bar: 50 µm.
Figure 12.
 
Cross-sectional H&E staining and immunostaining of collagen IV in normal cornea and endothelium wound cornea after 28 days of organ culture. (A) Intact DM was on the posterior surface of the cornea. (B) Multilayers of epithelial cells were observed on the anterior surface of the cornea. (C) An endothelium wound was created by peeling the central 8.5 mm DM. (D) The corneal epithelium in the wound cornea was thinner than that of the normal cornea. Scale bar: 100 µm. (EG) Normal cornea showed intact collagen IV expressions on DM across the peripheral to central cornea, with a single cell layer of endothelial cells. (IK) A central 8.5 mm cornea defect in DM was confirmed by labelling collagen IV of DM. Cells were found on the wound area, residing in the posterior stroma. (H, L) Images showed collagen IV expression on the epithelium's basement membrane in normal and wound corneas. Also, the thickness of the epithelium in the wound cornea was thinner than in the normal cornea. Representative images, n = 3 biological replicates. Cell nuclei were counterstained with Hoechst. Scale bar: 50 µm.
Table.
 
Summary of Donor Information
Table.
 
Summary of Donor Information
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×