Reduced Retinal Function in the Absence of Na(v)1.6

Abstract

Background: Mice with a function-blocking mutation in the Scn8a gene that encodes Na(v)1.6, a voltage-gated sodium channel (VGSC) isoform normally found in several types of retinal neurons, have previously been found to display a profoundly abnormal dark adapted flash electroretinogram. However the retinal function of these mice in light adapted conditions has not been studied.

Methodology/principal findings: In the present report we reveal that during light adaptation these animals are shown to have electroretinograms with significant decreases in the amplitude of the a- and b-waves. The percent decrease in the a- and b-waves substantially exceeds the acute effect of VGSC block by tetrodotoxin in control littermates. Intravitreal injection of CoCl(2) or CNQX to isolate the a-wave contributions of the photoreceptors in littermates revealed that at high background luminance the cone-isolated component of the a-wave is of the same amplitude as the a-wave of mutants.

Conclusions/significance: Our results indicate that Scn8a mutant mice have reduced function in both rod and the cone retinal pathways. The extent of the reduction in the cone pathway, as quantified using the ERG b-wave, exceeds the reduction seen in control littermates after application of TTX, suggesting that a defect in cone photoreceptors contributes to the reduction. Unless the postreceptoral component of the a-wave is increased in Scn8a mutant mice, the reduction in the b-wave is larger than can be accounted for by reduced photoreceptor function alone. Our data suggests that the reduction in the light adapted ERG of Scn8a mutant mice is caused by a combination of reduced cone photoreceptor function and reduced depolarization of cone ON bipolar cells. This raises the possibility that Na(v)1.6 augments signaling in cone bipolar cells.

Figure 1. The a- and b-waves of Scn8a ^dmu mice is reduced under nearly all stimulus and background conditions.

(a) Intensity series comparing representative Scn8a ^dmu (black) and control littermate (grey) ERGs over background luminances increasing from −0.5 to 1.9 log cd/m² (italicized) with a flash energies from −2.6 to 1 log cd·s/m². (b) Intensity response graphs for Scn8a ^dmu and control littermate a- and b-waves for backgrounds −0.5, 0.3, 0.6 0.9, 1.5, 1.9 log cd/m². (c) Isocline representation of the a-wave (top) and b-wave (bottom) in Scn8a ^dmu mice as the percent of control responses shown as a function of flash energy and background. Variance in isocline representation is equal to that of the intensity response curves in panel (b).

Figure 2. TTX does not affect the a-wave and the effect on the b-wave is most pronounced at lower stimulus intensity and background luminance.

(a) Intensity series comparing representative TTX-injected (black) and sham-injected control eye (grey) ERGs over background luminance increasing from −0.5 to 1.9 log cd/m² with a flash energies ranging from −2.6 to 1 log cd·s/m². (b) Intensity response graphs for TTX a and b-waves for backgrounds −0.5, 0.3, 0.6 log cd/m² (dim) and 0.9, 1.5, 1.9 log cd/m² (bright) (c) Isocline representation of the b-wave amplitude in TTX treated wild-type eyes mice as the percent of averaged control responses shown as a function of flash energy and background. Variance in isocline representation is equal to that of the intensity response curves in panel (b).

Figure 3. At high background luminances, the amplitude of the cone-isolated component of the a-wave of controls is similar to the a-wave of Scn8a ^dmu mice.

(a) Intensity series comparing a-wave amplitudes of Scn8a ^dmu mice (black) to control a-waves (grey) and control a-waves in the presence of either CNQX (red) or cobalt (blue) for dark adapted (DA) conditions. (b) Intensity series in light adapting conditions. (c) Intensity response graphs for all four conditions in dim (−0.5 and 0.3 log cd/m²), medium (0.6 and 0.9 log cd/m²) and bright (1.5 and 1.9 log cd/m²) backgrounds (italicized). (d) Isocline representation of the CNQX treated a-wave in as the percent of control responses shown as a function of flash energy and background. Variance in isocline representation is equal to that of the intensity response curves in panel (b).

Figure 4. Cones are present at normal density in Scn8a ^dmu mice as indicated by FITC-conjugated peanut agglutinin lectin staining in Scn8a ^dmu homozygous mice (dmu) and wild-type littermates (control).

Bar = 10 µm.