Inducing dry eye disease using a custom engineered desiccation system: Impact on the ocular surface including...

Dry eye disease (DED) is characterized by loss of tear film stability that becomes self-sustaining in a vicious cycle of pathophysiological events. Currently, desiccation stress (DS) is the dominant procedure for inducing DED in mice, however its' effect on limbal epithelial stem cells (LESCs) has been overlooked. This study aimed to establish a DS model via the use of a novel hardware to investigate the impact on the ocular surface including LESCs.

A mouse transporter unit was customized to generate a dehumidified environment. C57BL/6J mice were exposed to mild DS and injected with scopolamine hydrobromide (SH) or remained untreated (UT) under standard vivarium conditions for 10 consecutive days (n = 28/group). Clinical assessments included phenol red tear-thread test, fluorescein staining and optical coherence tomography assessments. Histopathological and immunofluorescence was used to evaluate tissue architecture, goblet cell (GC) status, lacrimal gland (LG) inflammation and epithelial phenotype on the ocular surface. Whole flat-mounted corneas were immunostained for keratin-14 (K14), then imaged by confocal microscopy and analyzed computationally to investigate the effect of DS on LESCs.

Custom modifications made to the animal transporter unit resulted in dehumidified cage relative humidity (RH) of 43.5 ± 4.79% compared to the vivarium 53.9 ± 1.8% (p = 0.0243). Under these conditions, aqueous tear production in mice was suppressed whilst corneal permeability and corneal irregularity significantly increased. H&E staining indicated stressed corneal basal epithelial cells and increased desquamation. DS-exposed mice had reduced GC density (41.0 ± 5.10 GC/mm vs 46.9 ± 3.88 GC/mm, p = 0.0482) and LGs from these mice exhibited elevated CD4⁺ cell infiltration compared to controls. DS elicited K14⁺ epithelial cell displacement, as indicated by increased fluorescence signal at a distance of 50-100 μm radially inwards from the limbus [0.63 ± 0.053% (DS) vs 0.54 ± 0.060% (UT), p = 0.0317].

Application of mild DS using customized hardware and SH injections generated features of DED in mice. Following DS, ocular surface epithelial cell health decreased and LESCs appeared stressed. This suggested that potential downstream effects of DS on corneal homeostasis are present, a phenomenon that is currently under-investigated. The method used to induce DED in this study enables the development of a chronic model which more closely resembles disease seen in the clinic.