Corneal wound healing is a complex, multilayered process initiated by injuries such as chemical exposure (Clippinger et al., 2021). This process comprises epithelial, stromal and endothelial layers, depending on the depth of the injury. The cells in these layers share common mechanisms in wound healing, such as cell migration and proliferation, the involvement of growth factors and cytokines, and extracellular matrix reorganization (Saghizadeh et al., 2017). In particular, the ocular surface relies on limbal epithelial/progenitor stem cells (LEPC) to renew the protective corneal epithelium. This cells reside in the specialized limbal transition zone, and are supported by niche cells like limbal mesenchymal stromal cells (LMSC), which maintain homeostasis and have direct injury-repair roles (Saghizadeh et al., 2017).

Stromal stem cells were first identified in the transitional zone between the cornea and sclera, known as the limbus, through immunostaining for ABCG2 (ATP-binding cassette transporter) (Hovakimyan et al., 2012; Singh et al., 2021). These cells are localized in the anterior stroma beneath the epithelial basement membrane, especially in regions with characteristic folds known as the palisades of Vogt (Du et al., 2005). These unique anatomical features, serve as a niche for LEPC (Polisetti et al., 2022a, Polisetti et al., 2022b; Tseng, 1996). Tomasello et al. (2016) also identified a subset of limbal stem cells expressing ABCG2, which were classified as mesenchymal stem cells (MSC) based on their expression of stem cell surface markers, including SSEA4 (stage-specific embryonic antigen-4), as well as nuclear transcription factors such as OCT4 (octamer-binding transcription factor 4), NANOG (homeobox protein NANOG), and SOX2 (SRY-box transcription factor 2), which are essential for self-renewal and the maintenance of pluripotency in both embryonic and adult stem cells (K. Takahashi et al., 2007; Yu et al., 2007).

The presence of MSC in the limbal stroma has been confirmed by multiple independent studies (Branch et al., 2012; Hashmani et al., 2013; G. Li et al., 2018; G. G. Li et al., 2012; Polisetti et al., 2022a; Sahoo et al., 2023; Sidney et al., 2015). Branch et al. (2012) has reported that such stromal stem cells from human limbus were multipotent and fulfilled the criteria for MSC developed by the International Society for Cellular Therapy's (ISCT). According to this committee, MSC are characterized by the expression of the surface markers CD105, CD73, and CD90, while lacking expression of CD45, CD34, CD14, CD19, and HLA-DR7 (Dominici et al., 2006).

Testing the ocular toxicity potential of chemicals is globally needed for product registration and labeling. Historically, the Draize rabbit test has been the standard for categorizing irritation severity based on induced lesions and reversibility time frame (OECD 405, 2023On the basis of the application of this test, the following ocular damage categories were established according to the United Nations Global Harmonized System (UN GHS): Category 1 (irreversible damage), Category 2 A/2B (reversible damage), and No Category (no damage or reversible within 24 h) (Luechtefeld et al., 2016). However, the Draize test has been widely criticized for its low reproducibility, subjective scoring system, anatomical differences between rabbit and human eyes, and ethical concerns, all of which have driven the search for alternative methods (Adriaens et al., 2014; Clippinger et al., 2021; Luechtefeld et al., 2016).

For over two decades, researchers have sought to replace the Draize eye test with in vitro alternatives, yet only partial success has been achieved to date (Adriaens et al., 2014). These alternatives include organotypic ex vivo assays, such as the Bovine Corneal Opacity and Permeability (BCOP) test (OECD 437, 2023), tissue engineered constructs such as the Reconstructed Human Cornea-like Epithelium (RhCE) (OECD 492, 2024) and in vitro cytotoxicity cell-based models such as the Short Time Exposure (STE) test (OECD 491, 2023), amongst others (Clippinger et al., 2021). While existing alternative methods can effectively analyze non-irritancy and severe damage, they cannot reliably determine moderate, reversible effects or recovery time frames needed for UN GHS classification (Adriaens et al., 2014).

Understanding the reversibility of ocular damage is crucial for improving current in vitro models for chemical hazard assessment. Existing organotypic methods, such as the BCOP (OECD 437, 2023), are effective in detecting immediate severe effects but fail to predict the persistence of damage, leading to frequent under-predictions of chemicals classified in vivo as Category 1 primarily based on the persistence of effects (Adriaens et al., 2014). Organotypic models such as the Ex-Vivo Eye Irritation Test (EVEIT) (Schrage et al., 2012; Spöler et al., 2015) and the Porcine Corneal Ocular Reversibility Assay (PorCORA) (Donahue et al., 2011; Piehl et al., 2010; Vij et al., 2024) have attempted to address this gap by tracking the healing process in excised corneas cultured over several days. However, neither method has undergone formal validation, emphasizing the necessity for further development and standardization of models that incorporate damage reversibility into hazard classification (Adriaens et al., 2014). In this context, the development and evaluation of a novel in vitro model capable of predicting the reversibility of damage offers a significant opportunity to improve the accuracy of chemical classification while advancing efforts to replace in vivo methods. To address these gaps, this study explores an alternative in vitro approach using bovine-derived LMSC.

Cattle are a key agricultural species, and the bovine cornea is one of the byproducts of the meat industry, which is utilized in the BCOP test (OECD 437, 2023). Previous research has shown that primary stromal cells can be isolated from fresh bovine corneas, with subsequent analysis confirming the expression of stem cell markers such as ABCG2, CD90, and CD73 (Funderburgh et al., 2005). As previously mentioned, LMSC play a crucial role in corneal regeneration. Given the accessibility of bovine corneas, this study aimed to isolate LMSC from this tissue and evaluate their potential for assessing the effects of reference substances categorized by the UN GHS for ocular irritation with particular emphasis on damage reversibility.