Monitoring Refractive Development in Mice to Investigate Mechanisms of Myopia

To reduce variability, measurements should be taken from an optically clear, bright, and dilated eye. Finding the location of the optic disk, noted by a significant reduction in image brightness, will aid in alignment and allow for rotation of the eye to a location just outside of this region. These locations will usually result in on-axis alignment, provide similar measurements in different cardinal directions, and result in low standard deviations. This process is made significantly easier in anesthetized mice.

Due to variability, a refractive difference of at least 2-3 diopters would probably be robust enough to resolve. If the eyes are optically clear, bright, and have been aligned well, measurement variability should be reduced. However, even at baseline some cohorts of mice can be more hyperopic than others (by several diopters). Therefore, in addition to comparing raw refractive error, comparing the change from baseline may also be useful. Generally, we try to have sample sizes of at least 10 mice, but this may depend on the specific experiment.

We have not readily seen an effect of anesthesia on refractive error measurements but there are published reports that typical anesthetics used in mice can induce hyperopic refractive errors, meaning that mice are myopic at young ages rather than hyperopic. Regardless of the absolute value, it does appear that mice became more hyperopic relative to their starting value as they age.

Ambient luminance is indeed important in myopia development. Form-deprivation myopia using diffuser lenses placed over the eye may reduce light transmission to the eye more than lens-induced myopia which uses powered lenses to defocus the image without greatly obstructing the passage of light. While there may be a small amount of luminance reduction with diffusion or defocus, it would not be expected to affect refractive development in pronounced way. This expectation is derived from studies demonstrating that changes in luminance (e.g. several orders of magnitude) do not affect refractive development in naïve eyes (no lens defocus).

This criterion is usually only applied at the baseline measurement, before any mice are treated with lenses. If the two eyes differ by more than 2 diopters at baseline, then it would be difficult to interpret any potential myopic shift in treated mice. Thus, we use this as exclusion criteria at the start. Naïve mice (those without treatment) with similar refractive error in both eyes at baseline usually do not, and should not, have pronounced differences in refractive error between their eyes as they age.

Yes, we have noticed that some mouse strains or genetic mutants can be more difficult to measure, possibly due to differences in optical quality or tear film.

The retinoscleral pathway that controls refractive development and myopia is unclear and remains a topic of active investigation by us and others. Rather than retina to sclera projections through brain centers, it is currently thought that retina to sclera signaling cascades occur locally within the eye. This is because eyes with severed optic nerves will still become myopic. In this case, signals originating in the retina would need to cross the RPE and choroid (or be translated to other signals in these tissues) to reach the sclera. Some promising candidate signals that are under investigation include dopamine and all-trans retinoic acid.