Effects of potent inhibitors of the retinoid cycle on visual function and photoreceptor protection from light damage in mice

Abstract

Regeneration of the chromophore 11-cis-retinal is essential for the generation of light-sensitive visual pigments in the vertebrate retina. A deficiency in 11-cis-retinal production leads to congenital blindness in humans; however, a buildup of the photoisomerized chromophore can also be detrimental. Such is the case when the photoisomerized all-trans-retinal is produced but cannot be efficiently cleared from the internal membrane of the outer segment discs. Sustained increase of all-trans-retinal can lead to the formation of toxic condensation products in the eye. Thus, there is a need for potent, selective inhibitors that can regulate the flux of retinoids through the metabolism pathway termed the visual (retinoid) cycle. Here we systematically study the effects of the most potent inhibitor of this cycle, retinylamine (Ret-NH2), on visual function in mice. Prolonged, sustainable, but reversible suppression of the visual function was observed by Ret-NH2 as a result of its storage in a prodrug form, N-retinylamides. Direct comparison of other inhibitors such as fenretinide and 13-cis-retinoic acid showed multiple advantages of Ret-NH2 and its amides, including a higher potency, specificity, and lower transcription activation. Our results also revealed that mice treated with Ret-NH2 were completely resistant to the light-induced retina damage. As an experimental tool, Ret-NH2 allows the replacement of the native chromophore with synthetic analogs in wild-type mice to better understand the function of the chromophore in the activation of rhodopsin and its metabolism through the retinoid cycle.

Fig. 1

Effects of different doses of Ret-NH₂ on the 11-cis-retinal and all-trans-retinyl ester levels in the eye. The 48-h dark-adapted C57BL/6 mice were gavaged with a single dose of Ret-NH₂ (1.75, 3.5, or 17.5 µmol). Next, 24 h after gavage, the mice were exposed to background light at 150 cd/m² for 20 min. Retinoid levels, 11-cis-retinal (A) and all-trans-retinyl ester (B), in the eye were examined at various times of dark adaptation. Retinoids were analyzed using HPLC methods as described under Materials and Methods. Inset, subset of mice gavaged with 3.5 µmol of Ret-NH₂ were exposed a second time to the same intensity of illumination 7 days after the first exposure to light. All the procedures were carried out in the dark. The gray bar indicates the levels of all-trans-retinyl esters in wild-type mice. Mean ± S.D. was indicated (n= 3 for each point).

Fig. 2

Clearance of Ret-NH₂ in the liver, blood, and eye. The 48-h dark-adapted C57BL/6 mice were gavaged with a single dose of Ret-NH₂ (1.75, 3.5, or 17.5 µmol). After gavage, the levels of Ret-NH₂ and N-retinylpalmitamide were measured in the liver (A, D), blood (B, E), and eye (C, F) at various time points. Retinoids were analyzed using HPLC methods as described under Materials and Methods. Mean ± S.D. was indicated (n= 3 for each point).

Fig. 3

Comparison of the effect of Ret-NH₂ and other inhibitors on the retinoid cycle. The 48-h dark-adapted C57BL/6 mice were gavaged with a single dose of N-retinylamides, Ret-NH₂, fenretinide, or 13-cis-retinoic acid (0.35,1.75, or 3.5 µmol). The mice were exposed to background light at 150 cd/m² for 20 min at 24 h after gavage. The 11-cis-retinal in the eye and Ret-NH₂ in the liver were analyzed at 5 h (A; eye) (C; liver) and 24 h (B; eye) (D; liver) after the bleach. Retinoids were analyzed using HPLC methods as described under Materials and Methods. Mean ± S.D. was indicated (n= 3 for each point).

Fig. 4

Gene array analysis. The expression levels of mRNA were compared between mice treated with 13-cis-retinoic acid, fenretinide, and Ret-NH₂ using a 37,364 cDNA array (provided by NimbleGen System Inc.) as described under Materials and Methods. Normalized values of mRNA expression were plotted, control versus each treated group for liver mRNA (top) and for eye mRNA (bottom) as scattered plot with Sigma Plot v.9.0. Numbers of genes for which expression level was changed to less than 0.5 of control (dark blue points) or to more than 2 after the treatment (red points) were indicated, whereas the levels of expression within the experimental error is shown in light blue.

Fig. 5

Deficiencies of arrestin and transducin a-subunit translocation in Ret-NH₂-treated mice. Left column indicates the mice without Ret-NH₂ administration, and right column shows the mice gavaged with Ret-NH₂. A, immunofluorescent localization of rod arrestin (red). In the dark adapted retinas, arrestin was distributed throughout the retina. Under light-adapted conditions, translocation of arrestin to photoreceptor OS was observed in control mice (left, bottom). Only minimal translocation of arrestin to OS was observed in light-adapted mice gavaged with Ret-NH₂. B, immunofluorescent localization of transducin α-subunit (red). In the dark-adapted retinas, transducin α is localized in photoreceptor OS. In light-adapted retinas of control mice (bottom left), transducin is localized in IS, ONL, and OPL. This light-dependent translocation is mostly abolished in mice gavaged with Ret-NH₂, which shows the transducin α immunoreactivity mostly in OS and weakly in OPL. IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Nuclei are stained with Hoechst 33342 (blue).

Fig. 6

Substitution of 11-cis-retinal by 9-cis-retinal. The 48-h dark-adapted C57BL/6 mice were gavaged once with Ret-NH₂ (3.5 or 17.5 µmol), and 16 h after gavage the mice were exposed to background light at 500 cd/m² for 24 min. Next, mice were placed in the dark and 2 h after exposure to light gavaged with 9-cis-retinyl acetate (7.0 µmol) and kept in the dark before the retinoid analysis. 9-cis-Retinal (A, top) and 11-cis-retinal (A, bottom) levels were analyzed by HPLC as described under Materials and Methods. Mean ± S.D. was indicated (n= 3 for each point). Two days after 9-cis-retinyl acetate gavage, scotopic and photopic ERG was performed as described under Materials and Methods. Three mice from each group were examined, and representative scotopic waves were indicated (B).

Fig. 7

The protective effect of Ret-NH₂ on light-induced retinal damage in a mouse model. The 48-h dark-adapted BALB/c mice were gavaged once with Ret-NH₂ (3.5 µmol). After 16 h, mice were exposed to 5000 lux fluorescent light for 2 h to induce light damage of the retina. For the remaining time, mice were kept in the dark. A, morphology of the retina was examined 7 days after the light exposure. Left, the retina section from the control mouse not exposed to the intense light; middle, the retina section from mouse exposed to the intense light but treated with Ret-NH₂; right, the retina section from the control mouse exposed to the intense light treated with vehicle solution only. Similar results were obtained from four independent experiments. B, retinoid content of the eyes from mice treated and exposed to light as in A. The eyes were analyzed for retinoid content 2 weeks after the light exposure as described under Materials and Methods. The black arrows indicate all-trans-retinyl ester, and the open arrows indicate 11-cis-retinal in the chromatograms of retinoid analysis. C, rhodopsin concentration was measured 2 weeks after the light exposure. D, immunoblot of rhodopsin from the light damage-induced mouse eyes with or without Ret-NH₂ treatment. Rhodopsin monomer and dimer were detected in control and Ret-NH₂ pretreated mice, whereas aggregated rhodopsin was detected in the mice treated with vehicle solution. E through H, mice were dark-adapted for 48 h, and scotopic (E, F) and photopic (G, H) ERG were recorded 7 days after light exposure as described under Materials and Methods. Ret-NH₂-treated mice showed no significant differences compared with the control, whereas the amplitudes in mice treated with vehicle solutions were attenuated significantly compared with control and Ret-NH₂-treated mice (P < 0.0001). Statistical analysis was performed by one-way analysis of variance (n= 5 in each condition).