# V-Cornea: A computational model of corneal epithelium homeostasis, injury, and recovery

## TL;DR

V-Cornea is a computer model that simulates how the cornea heals after injury, offering a new way to study eye damage and recovery without animal testing.

## Contribution

V-Cornea introduces a novel agent-based computational model of corneal epithelium homeostasis and injury recovery, enabling mechanistic insights into healing processes.

## Key findings

- V-Cornea accurately simulates corneal epithelial architecture and homeostasis over extended periods.
- The model predicts healing timeframes of 3–5 days for superficial injuries and longer recovery for moderate injuries with basement membrane disruption.
- The simulation reproduces clinical patterns of corneal healing and instability resembling recurrent corneal erosion.

## Abstract

To develop a computational model that addresses limitations in current ocular irritation assessment methods, particularly regarding long-term effects, and recovery patterns following chemical exposure or trauma to the cornea.

V-Cornea is an agent-based computer simulation implemented in CompuCell3D that models corneal epithelial homeostasis and injury response. The model incorporates biologically-inspired rules governing cell behaviors (proliferation, differentiation, death) and key signaling pathways including Epidermal Growth Factor (EGF), translating cell-level behaviors to tissue-level outcomes (in vitro to in vivo extrapolation, IVIVE).

V-Cornea successfully reproduces corneal epithelial architecture and maintains tissue homeostasis over extended simulated periods. Following simulated trauma or toxicant exposure, the model accurately predicts healing timeframes of 3–5 days for slight and mild injuries. For moderate injuries with basement membrane disruption, the model demonstrates longer recovery times and emergent dynamic structural disorganization that mimics recurrent corneal erosions, providing mechanistic insights into these pathological conditions.

V-Cornea’s modular CompuCell3D implementation is easily extensible to incorporate additional recovery and injury mechanisms. Future versions will include more realistic basement membrane dynamics and explicit representation of stromal tissue and immune response, to improve predictivity for moderate injuries. This virtual-tissue approach shows potential not only for toxicological assessments but also for drug discovery and therapy optimization by providing a platform to test interventions and predict outcomes across various injury scenarios.

Injuries to the cornea are a common health concern, typically resulting from chemical exposure, physical trauma, infection, or environmental damage. Current methods for assessing eye irritation often rely on animal testing or simplified laboratory assays that provide only static snapshots of tissue damage. Clinical time series data on recovery are limited, making predicting long-term healing outcomes and distinguishing between superficial injuries (completely healed) versus deeper injuries (with persistent complications) difficult. To address these limitations in understanding and predicting corneal recovery, we developed V-Cornea, a computer simulation of the human corneal epithelium. By programming individual virtual cells with biologically-motivated rules for growth, movement, differentiation, and death, our model accurately builds and maintains the tissue’s layered structure and continuous 7–14 day renewal cycle. Crucially, the simulation mimics clinical healing patterns: superficial injuries repair fully within 3–5 days, while deeper injuries compromising the basement membrane lead to incomplete healing and instability resembling recurrent corneal erosion. This virtual tissue approach demonstrates how microscopic cellular mechanisms drive tissue-level recovery in the cornea. V-Cornea offers a flexible, animal-free platform for investigating chemical toxicity mechanisms and optimizing therapeutic interventions for corneal injuries, providing mechanistic insights into wound healing that are unattainable through traditional experimental assays.