Measurement of refractive state and deprivation myopia in two strains of mice

The mouse eye has a bright retinal image (f/number <1) but low optical quality (visual acuity about 0.5 cpd) that may render emmetropization unnecessary. However, this species is potentially a powerful model to study eye growth and myopia because its genome can be readily manipulated and has been completely sequenced. We have investigated how precisely eyes of mice can be refracted and tested whether deprivation myopia can be induced by frosted diffusers.

An automated eccentric infrared photorefractor was adapted to refract eyes of two mouse strains--C57BL/6 (B6) and DBA/2 (D2)--during Tropicamide cycloplegia without anesthesia. Axial lengths were measured in highly magnified video images of freshly excised eyes. Plastic hemispherical diffusers were applied between postnatal days and 29 and left attached for 7 or 14 days.

(1) Trial lenses ranging from +10 to -10 D produced high correlations between the brightness slope in the pupil and applied lens power (r = 0.81 and r = 0.87), demonstrating reliable refraction. Five repeated measures in 12 eyes showed an average standard deviation of 3.0 D, equivalent to an axial length change <10 microm (derived from schematic eye modeling). (2) Deprivation produced a significant shift toward myopia, relative to untreated eyes, but only after 14 days and only in B6 mice (p = 0.02 with or p = 0.00038 without one outlier; N = 9). In contrast, DBA/2J were unaffected by occlusion, perhaps due to mutations that target eye, lens, or anterior segment. (3) Both eyes of untreated animals often had axial lengths that differed markedly. Surprisingly, we detected no significant correlation between refractive error and axial length after treatment.

The infrared refraction technique is sufficiently sensitive to resolve equivalent changes in axial length of only +/- 10 microm in alert mice. Prolonged occlusion produces a significant myopic shift in B6 mice, but not in D2 mice. Even among isogenic B6 mice, the response is variable for reasons that presumably trace back to subtle developmental, environmental, and technical factors.