Validating the Mouse Model for Ocular Surface Reserach and Assessing Morphological Changes in the Dry Eye Palpebral Conjunctiva

Purpose: The young adult mouse has become the preferred animal model in many areas of ophthalmic research. For this reason, there is a need to validate its use in different experimental settings and to assess its applicability to the human. The purpose of this research was to firstly, evaluate the normal mouse cornea and conjunctiva, secondly to utilize the mouse as an animal model, to investigate structural changes occurring in the dry eye and thirdly to investigate ocular UVR transmittance of this small eye.

Methods: A total of 45 animals were utilized for these experiments. Ocular dryness was induced in 16 by administration of scopolamine hydrobromide (0.5mg/0.2ml) QID, and exposure to a dry environment for 5 (DS5) or 10 (DS10) days. After euthanization the corneas or eye lids were fixed either in 2% glutaraldehyde and processed according to an established histology protocol for light- (LM) and transmission electron microscopy (TEM) or in 10% formalin and embedded in paraffin. Counts or measurements were obtained utilizing well defined specific reference points allowing the observer to conduct accurate and repeatable intra-corneal or intra-conjunctival measurements. Goblet cell density was investigated both on resin and paraffin embedded tissues with a variety stains and histological protocols. UVR transmittance of the whole mouse eye, the cornea and the crystalline lens was also investigated. Immediately after euthanization the mouse eye was placed under a 0.2 mm diameter detecting fiber optic illuminated by an Oceans Optics deuterium/halogen source (DT-MINI-2-GS). The data were recorded and analyzed with an Ocean Optics spectrometer (USB4000) and analyzed in Matlab. Results: The mouse epithelium contributed to 1/3 and the stroma to 2/3 of the total corneal thickness. A statistically significant (P < 0.001) decrease in overall thickness was found in the periphery compared to the center. The central cornea had an average of 50 ± 2 lamellae with an average thickness of 2.1 µm versus 36 ± 3 and 1.8 µm peripherally (P < 0.005). The mouse did not possess an anterior limiting lamina (ALL).
Near the junction between the lid margin and the normal palpebral conjunctiva the epithelium had an average thickness of 46 ± 11 µm and 9 ± 2 cell layers, versus 42 ± 7 µm and 8 ± 1 layers in the DS5 (P > 0.05) and 38 ± 6 µm and 7 ± 1 layers in the DS10 (P < 0.05). In the goblet cell populated palpebral region the normal conjunctival epithelium averaged 24 ± 4 µm and was significantly (P < 0.05) thicker than in the DS5 (18 ± 3 µm) and the DS10 (19 ± 3 µm). In the control 43% of the goblet cells had epithelial coverage, compared to 58% for the DS5 and 63% for the DS10 (P < 0.05). Some stains indicated a decreased number of goblet cells in the dry eye (PAS and Alcian blue), while (MUC5AC, histology) did not. The mouse had a UVR transmittance cut off at 280 nm for the cornea and 310 nm for the crystalline lens. 50% absorption of UVR was achieved at 330 nm and 375 nm by the cornea and crystalline lens respectively. The combined transmittance spectra of the cornea and the lens compared well with the total UVR transmittance for the whole eye and, thus, supported the validity of our findings.

Conclusions: The mouse cornea becomes significantly thinner towards the periphery and this was explained by a decreased thickness of both the epithelium and the stroma. Stromal thinning was caused by a reduction in number of lamellae towards the periphery. This finding suggests that not all lamellae cross the cornea limbus to limbus. A notable difference between the mouse and the human cornea is the absence of an ALL in the former. The palpebral conjunctival epithelium decreased in thickness in the dry eye and the goblet cell access to the surface appears to be inhibited by surrounding epithelial cells, which potentially may slow down their migration to the surface. Goblet cell density differed depending on stain used for evaluation of mucous secreting cells, suggesting that there may be different subtypes of goblet cells on the ocular surface or may be PAS and Alcian blue are mucin specific and not goblet cell specific. It is important to be aware of the ultrastructural and dimensional differences between the mouse and the human cornea when utilizing the mouse as a model. The mouse conjunctival epithelium has a structure similar to the human and appears to be viable model for conjunctival and dry eye research. However, due to the differences between the mouse and human in UVR transmittance, it is concluded that the mouse is not an ideal model for UVR experiments.