Mutations in the gene encoding ABCR are responsible for Stargardt macular dystrophy. Here we show by immunofluorescence microscopy and western-blot analysis that ABCR is present in foveal and peripheral cone, as well as rod, photoreceptors. Our results suggest that the loss in central vision experienced by Stargardt patients arises directly from ABCR-mediated foveal cone degeneration.
Macular degeneration is a heterogeneous group of common retinal disorders characterized by a progressive loss in central vision. Mutations in the gene (ABCA4) encoding the photoreceptor-specific, ATP-binding cassette transporter ABCR are responsible for disease variants including Stargardt macular dystrophy1, cone-rod dystrophy2, RP-like dystrophy3,4 and age-related macular degeneration5 (although the latter remains controversial6). Affected individuals have decreased visual acuity and progressive bilateral atrophy of the cone-rich fovea and underlying retinal pigment epithelium (RPE). Therefore, it was an unexpected result that several groups found ABCR in rod, but not cone, photoreceptors1,7,8.
The fovea is a small, specialized region of central retina or macula responsible for high visual acuity in humans. The centre of the fovea, known as the foveal pit, has a decreased thickness due to the outward displacement of inner retinal neurons, maximum cone density and an absence of rod photoreceptors9. To determine if ABCR is present in cones, we double-labelled foveal cryosections overnight with antibodies specific for ABCR (ref. 8) and green/red cone-opsin for cones7, or rhodopsin for rods. Anti-ABCR antibodies specifically labelled the photoreceptor outer segment layer (Fig. 1a,c), extending from the foveal rim (containing cones and rods) through the cone-rich foveal pit (Fig. 1b), which is devoid of rods ( Fig. 1d). We used four antibodies specific for distinct epitopes on ABCR (including antibodies used in previous studies7,8) and all showed the same labelling pattern. We also determined the presence and size of ABCR in foveal extracts by SDS gel electrophoresis and western blot. ABCR migrated as a single 220-kD protein in extracts derived from fovea, devoid of detectable rods; macula, containing both rods and cones; and peripheral retina, containing predominantly rods (Fig. 1e). These results suggest that the same gene encodes ABCR in rod and cone photoreceptors.
To determine if ABCR is present in peripheral cones, we labelled retinal whole mounts with antibodies against ABCR and green/red cone opsin, blue cone opsin or rhodopsin. Anti-ABCR antibodies labelled the mosaic of green/red cone (Fig. 2a–d) and less numerous blue cone (Fig. 2e,f) outer segments among the numerous labelled rod outer segments, contrary to an earlier report7. As the same antibodies were used in our and previous studies, the detection of ABCR in peripheral cones demonstrated here is probably due to the labelling conditions and tissue samples used to detect ABCR in cones among the more numerous rods. Our findings are consistent with an earlier study showing that the rim protein, the frog orthologue of ABCR, is present in rods and cones10.
Individuals with Stargardt disease and other ABCR-linked retinopathies typically show symptoms related to cone dysfunction. In addition to reduced visual acuity, affected individuals often experience abnormal colour vision and have decreased cone electroretinograms11,12 (ERGs). Loss of cone function was initially thought to be a secondary consequence of rod-induced dysfunction of RPE cells1. Some Stargardt patients, however, have diminished cone ERGs with normal rod ERGs, leading one to question the primary involvement of rods in cone dysfunction12. ABCR expression in cones indicates a primary function of cones in ABCR-linked diseases. We suggest that decreased visual acuity and atrophy of the central retina is a direct consequence of ABCR-mediated foveal cone degeneration, whereas reduced full-field cone ERGs observed in some Stargardt patients result from ABCR-mediated peripheral cone degeneration. Phenotypic variations in Stargardt patients may be partially due to differential effects of different ABCR mutants on cone and rod function and viability.
Several lines of evidence now suggest that ABCR functions as a retinoid transporter to facilitate the removal of all-trans retinal from rod outer segment disks after the photobleaching of rhodopsin. The ATPase activity of purified, reconstituted ABCR is stimulated up to fivefold by retinal13, and mice homozygous for disrupted Abca4 accumulate retinoid compounds in rods and RPE (ref. 14). Photobleaching of cone opsin also produces all-trans retinal. We suggest that ABCR transports retinal across cone disk membranes to facilitate its reduction to retinol by retinol dehydrogenase. Disease-linked mutations in ABCR diminish this activity, leading to the accumulation of retinal derivatives in cone disk membranes and the diretinal compound A2E in underlying phagocytic RPE cells15. Foveal RPE may be particularly sensitive to A2E accumulation, leading to preferential degeneration of foveal RPE and cone photoreceptors and loss of central vision as an early symptom of these diseases.
Allikmets, R. et al. Nature Genet. 15, 236– 246 (1997).
Cremers, F.P.M. et al. Hum. Mol. Genet. 7, 355– 362 (1998).
Martinez-Mir, A. et al. Nature Genet. 18, 11– 12 (1998).
Klevering, B.J. et al. Br. J. Ophthalmol. 83, 914– 918 (1999).
Allikmets, R. et al. Science 277, 1805– 1807 (1997).
Stone, E.M. et al. Nature Genet. 20, 328– 329 (1998).
Sun, H. & Nathans, J. Nature Genet. 17, 15–16 (1997).
Illing, M., Molday, L.L. & Molday, R.S. J. Biol. Chem. 272, 10303– 10310 (1997).
Osterberg, G. Acta Ophthalmol. 6, 1–103 (1935).
Papermaster, D.S., Reilly, P. & Schneider, B.G. Vision Res. 22, 1417– 1428 (1982).
Fishman, G.A. et al. Arch. Ophthalmol. 117, 504– 510 (1999).
Lois, N., Holder, G.E., Fitzke, F.W., Plant, C. & Bird, A.C. Invest. Ophthalmol. Vis. Sci. 40, 2668–2675 (1999).
Sun, H., Molday, R.S. & Nathans, J. J. Biol. Chem. 274, 8269– 8281 (1999).
Weng, J. et al. Cell 98, 13–23 (1999).
Eldred, G.E. & Lasky, M.R. Nature 361, 724–726 (1993).
We thank J. Nathans for providing the ABCR1156-1258, ABCR425-570, green/red cone opsin JH 492 and blue cone opsin JH 455 polyclonal antibodies; J. Hurley for the cone transducin A1.1 antibody; A. Bird for discussions; and the UBC Eye Bank for human donor eyes. Support was provided by grants from the Ruth and Milton Steinbach Foundation, National Eye Institute and the Medical Research Council of Canada.
About this article
Japanese Journal of Ophthalmology (2012)