Cone-based vision in the aging mouse

Abstract

People often experience age-related declines in cone-based visual capacities despite an absence of apparent visual pathology. Although mice are used as models of human visual pathologies associated with aging, little is known about how age impacts vision in animals with disease-free retinas since most studies have heretofore examined relatively young mice. We examined the effects of age on cone-based vision by assessing opsin gene transcription, cone densities, the flicker electroretinogram (ERG), and behavioral increment thresholds in mice. ERG measurements of cone function showed age-related declines in maximum voltage (Vmax), while opsin gene transcription, cone density, and increment thresholds were unchanged even in extremely old mice. The age-related decline in Vmax seen in mice is qualitatively similar to that documented for human subjects. It is notable that Vmax, a commonly used index of ERG activity, does not predict behavioral performance in the mouse.

Fig. 1

Mouse M opsin (circles, solid line) and UV opsin (squares, dashed line) mRNA transcript levels as a function of age. The lines are best-fitting linear regressions. The mRNA transcript levels were normalized to the geometric mean of a suite of housekeeping genes and are plotted on a logarithmic scale.

Fig. 2

Immunolabeled flat-mounted mouse retinas. Shown are representative images of retinas from two young (ages = 105 and 185 days) mice and one old mouse (age = 857 days) double-labeled with PNA (middle column) and a cone opsin-specific antibody (left column). The right column shows images of opsin antibody-labeled cells (left column) superimposed with cells labeled with PNA (middle column). In all cases opsin antibody reactivity was restricted to cone outer segments. Scale bar = 20 μM.

Fig. 3

Results from fluorescence immunohistochemistry experiments. (a) Estimated cone densities (cells/mm²) as evidenced by PNA reactivity. The densities are plotted as a function of age (n = 7) with the best-fitting line. (b) Average densities (cells/mm²) of cones expressing M pigment (triangles; n = 5) and UV pigment (circles; n = 5) plotted as a function of age. (M pigment, solid line; UV pigment, dashed line).

Fig. 4

Representative V–log I functions from a single young mouse (circles) and a single old mouse (squares). The stimulus was a flickering light with a wavelength content designed to be absorbed by the M pigment (see text for complete details). Each data set has been fit using the Naka–Rushton equation.

Fig. 5

Parameters derived from the best-fits to V–log I functions obtained from 30 mice tested at different ages. The stimulus was a flickering light having wavelength content such that it activated receptors containing M cone pigment. (a) V_max values. The dashed lines show the bounds of the 99% confidence interval. (b) k parameter. (c) Slope parameter.

Fig. 6

Parameters derived from the best-fits to V–log I functions obtained from 30 mice tested at different ages. The stimulus was designed to activate cones containing UV pigment. Other details as for Fig. 5.

Fig. 7

(a) Two-point increment threshold spectral sensitivity functions measured under photopic (open circles) and scotopic (closed circles) conditions. The photopic data were best-fit (dashed lines) by a photopigment absorption function having a peak of 509 nm. The best-fit curve for scotopic data (continuous line) has a peak of 500 nm. (b) Behavioral increment thresholds measured under photopic conditions from a single mouse over a span of 21 months (107–753 days; open circles) with the best-fitting linear regression (solid line). Also shown are thresholds from three geriatric mice measured at various ages (662–905 days; open triangles and diamonds) and the average threshold from three young mice (108 days of age; open square, ±SD). (c) Behavioral increment thresholds measured under scotopic conditions from a single mouse over a span of ∼24 months (89–801 days; closed circles) with the best-fitting line (solid line). Also shown are scotopic increment thresholds obtained from a second geriatric mouse (935 days; closed triangle) and the average threshold obtained from three young mice (87–217 days; closed square, ±SD). All thresholds were derived as described in the text.