Generation and Feasibility Assessment of a New Vehicle for Cell-Based Therapy for Treating Corneal Endothelial Dysfunction

Abstract

The corneal endothelium maintains corneal transparency by its pump and barrier functions; consequently, its decompensation due to any pathological reason causes severe vision loss due to corneal haziness. Corneal transplantation is the only therapeutic choice for treating corneal endothelial dysfunction, but associated problems, such as a shortages of donor corneas, the difficulty of the surgical procedure, and graft failure, still need to be resolved. Regenerative medicine is attractive to researchers as a means of providing innovative therapies for corneal endothelial dysfunction, as it now does for other diseases. We previously demonstrated the successful regeneration of corneal endothelium in animal models by injecting cultured corneal endothelial cells (CECs) in combination with a Rho kinase (ROCK) inhibitor. The purpose of the present study was to optimize the vehicle for clinical use in cell-based therapy. Our screening of cell culture media revealed that RELAR medium promoted CEC adhesion. We then modified RELAR medium by removing hormones, growth factors, and potentially toxic materials to generate a cell therapy vehicle (CTV) composed of amino acid, salts, glucose, and vitamins. Injection of CECs in CTV enabled efficient engraftment and regeneration of the corneal endothelium in the rabbit corneal endothelial dysfunction model, with restoration of a transparent cornea. The CECs retained >85% viability after a 24 hour preservation as a cell suspension in CTV at 4°C and maintained their potency to regenerate the corneal endothelium in vivo. The vehicle developed here is clinically applicable for cell-based therapy aimed at treating the corneal endothelium. Our strategy involves the generation of vehicle from a culture medium appropriate for a given cell type by removing materials that are not favorable for clinical use.

Fig 1. Generation of a human corneal endothelium cell (HCEC) vehicle for cell therapy.

(A) Effect of various media types on HCEC adhesion to the substrate. HCECs were seeded with media without supplementation with FBS, and then the numbers of adhering cell were evaluated after 24 hours. RELAR, M199, and F12/DMEM media significantly enhanced cell adhesion when compared to OptiMEM-I. *p<0.01. (B) The cell therapy vehicle (CTV) was generated based on RELAR medium but with removal of materials such as hormones, growth factors, and other materials with possible toxicity. The numbers of adhered cells increased by 120.7% with CTV when compared to OptiMEM-I, although cell adhesion tended to be lower than with RELAR. However, HCECs seeded with the Y-27632 ROCK inhibitor showed a similar level of cell adhesion in RELAR and CTV (135.6% and 131.1% vs OptiMEM-I without Y-27632, respectively). *p<0.01. (C) The numbers of adhered HCECs were significantly lower in liquid Opeguard-MA Intraocular Irrigation Solution (used for intraocular surgery) and Ringer’s solution (used for intravenous drip infusion) than with CTV. *p<0.01. (D) Cell viability of HCECs was evaluated by Trypan blue staining after preservation in CTV for 24 hours at 4°C. HCECs derived from 4 independent donors and preserved in CTV showed showed a greater than 85% exclusion of Trypan blue. All experiments were performed in at least triplicate.

Fig 2. Rabbit corneal endothelium cell (RCEC) injection with cell therapy vehicle (CTV) enables regeneration of corneal endothelium.

(A) A total of 5.0×10⁵ RCECs, suspended in 200 μl of intraocular irrigation solution, DMEM, or CTV supplemented with 100 μM Y-27632 was injected into the anterior chamber of the rabbit corneal endothelial dysfunction model (n = 3). CTV (200 μl) supplemented with 100 μM Y-27632 was injected into the anterior chamber of the rabbit corneal endothelial dysfunction model as a control (n = 3). Corneal transparency was restored by intracameral injection of RCECs suspended in either DMEM or CTV, while control eyes exhibited hazy corneas due to corneal endothelial dysfunction. (B) Central corneal thickness was evaluated with an ultrasound pachymeter and was restored to almost a normal value in the eyes transplanted with RCECs in DMEM or CTV. Eyes injected with RCECs in intraocular irrigation solution showed a thicker central corneal thickness when compared to eyes injected with RCECs in DMEM or CTV. (C) Intraocular pressure (IOP) elevation due to formation of cell aggregates in the eye is a possible complication and was evaluated with a Tonovet^®. The IOP remained in the normal range throughout the 2 weeks of the study in all groups. (D) Regenerated corneal endothelium was evaluated by immunofluorescent staining 2 weeks after cell transplantation. The function-related markers Na⁺/K⁺-ATPase (pump function), ZO-1 (tight junction), and N-cadherin (adherent junction) were expressed in all regenerated CECs in eyes from both the DMEM and CTV groups. Actin staining showed hexagonal regenerated corneal endothelial cells. By contrast, control eyes had few fibroblastic transformed cells and lacked expression of the function-related markers. Scale bar: 50 μm. (E) Cell density of regenerated corneal endothelium formed by injecting RCECs was the same for cells suspended in DMEM and CTV, while cells suspended in intraocular irrigation solution showed lower cell density (Fig 2E).

Fig 3. Rabbit corneal endothelium cells (RCECs) preserved in CTV regenerate corneal endothelium.

(A) Schematic image showing the protocol for assessment of the feasibility of cell preservation in CTV. RCECs were harvested from the culture plate and suspended in CTV for 24 hours at 4°C, the ROCK inhibitor, Y-27632, was added to the RCECs before injection, and RCECs were injected into the anterior chamber of the rabbit corneal endothelial dysfunction model (n = 3). As a control, CTV supplemented with Y-27632 was injected into the rabbit corneal endothelial dysfunction model (n = 3). (B) Slitlamp microscopy showed that a transparent cornea was formed by the preserved RCECs, while the control eyes injected with vehicle alone exhibited hazy corneas. (C) Scheimpflug images showed the restoration of an anatomically normal cornea similar to a healthy cornea following injection of preserved RCECs. Control eyes injected with vehicle only exhibited corneal edema due to corneal endothelial dysfunction. (D) Corneal thickness was evaluated with a Pentacam^®. The color map showing corneal thickness demonstrated that eyes injected with preserved RCECs exhibited normal thicknesses throughout the center to the periphery, whereas control eyes exhibited thick corneas after 2 weeks. (E) Corneal volumes evaluated with a Pentacam^® showed normal levels after RCEC injection. (F) Central corneal thickness evaluated with an ultrasound pachymeter showed that eyes injected with preserved RCECs had corneas of normal thickness. (G) Regenerated corneal endothelium was examined in vivo by contact specular microscopy. A hexagonal monolayer sheet structure was observed in the eyes injected with preserved RCECs, whereas no similar structure was observed in the control eyes.

Fig 4. Histological assessment of corneal endothelium regenerated by injection of RCECs and HCECs after preservation in CTV.

(A) Regenerated corneal endothelium was evaluated by immunofluorescent staining 2 weeks after cell transplantation. The function-related markers Na⁺/K⁺-ATPase, ZO-1, and N-cadherin were expressed in the corneal endothelium of the eyes injected with RCECs after 24 hours of preservation. Actin similar to that seen in healthy cells was observed along the cell cortex in the eyes injected with RCECs. Fibroblastic cells without function-related markers were observed. Scale bar: 50 μm. (B) Feasibility of using HCECs after preservation in CTV was evaluated by injecting HCECs preserved for 24 hours in CTV into the rabbit corneal endothelial dysfunction model (n = 3). Immunofluorescence staining showed that corneal endothelium was regenerated by the injected HCECs and expressed Na⁺/K⁺-ATPase and N-cadherin. Actin staining showed that the regenerated corneal endothelium in the rabbit eyes had a normal morphology consisting of a monolayer of hexagonal cells. Scale bar: 500 μm.