In vivo electroretinographic differentiation of rod, short-wavelength and long/medium-wavelength cone responses in dogs using silent substitution stimuli.

The canine species has dichromatic color vision comprising short-wavelength (S-) and long/medium (L/M-) wavelength-sensitive cones with peak spectral sensitivity of 429-435 nm and 555 nm respectively. Although differentiation of rod- and cone-mediated responses by electroretinogram (ERG) in dogs is commonly performed, and standards have been developed based on standards for human observers, methods to differentiate S- and L/M-cone responses in dogs have not been described. We developed flicker protocols derived from previously published rod and cone spectral sensitivities. We used a double silent substitution paradigm to isolate responses from each of the 3 photoreceptor subclasses. ERG responses were measured to sine-wave modulation of photoreceptor excitation at different temporal frequencies (between 4 and 56 Hz) and mean luminance (between 3.25 and 130 cd/m2) on 6 different normal dogs (3 adult female, and 3 adult male beagles) and one female beagle dog with suspected hereditary congenital stationary night blindness (CSNB). Peak rod driven response amplitudes were achieved with low frequency (4 Hz, maximal range 4-12 Hz) and low mean luminance (3.25 cd/m2). In contrast, peak L/M-cone driven response amplitudes were achieved with high frequency (32 Hz, maximal range 28-44 Hz) and high mean luminance (32.5-130 cd/m2). Maximal S-cone driven responses were obtained with low frequency stimuli (4 Hz, maximal range 4-12 Hz) and 32.5-130 cd/m2 mean luminance. The dog with CSNB had reduced rod- and S-cone-driven responses, but normal/supernormal L/M cone-driven responses. We have developed methods to differentiate rod, S- and L/M-cone function in dogs using silent substitution methods. The influence of temporal frequency and mean luminance on the ERGs originating in each photoreceptor type can now be studied independently. Dogs and humans have similar L/M cone responses, whereas mice have significantly different L/M responses. This work will facilitate a greater understanding of canine retinal electrophysiology and will complement the study of canine models of human hereditary photoreceptor disorders.