Analyses of Stargardt patients together with biochemical characterization of ABCA4 and abca4
knockout mice have implicated ABCA4 in removal of retinoids from outer segments as part of the visual or retinoid cycle [55
]. Following photoexcitation, all-trans
retinal is released from rhodopsin and partitions into the lipid bilayer of disc membranes. Most of the all-trans
retinal is reduced to all-trans
retinol on the cytoplasmic side of the membrane by the retinol dehydrogenase RDH8 and possibly other NADPH-dependent dehydrogenases [109
retinol is then transported from photoreceptors to the RPE cell where it is converted to all-trans
retinal ester by lecithin:retinol acyl transferase (LRAT), isomerized to 11-cis
retinol by RPE65-isomerase, and oxidized to 11-cis retinal by 11-cis retinol dehydrogenases (). 11-cis retinal is then transported back to the photoreceptor outer segment where is can recombine with opsin to regenerate rhodopsin or cone opsin [109
]. However, a fraction of all-trans
retinal generated from the photobleaching of rhodopsin or cone opsin reacts with PE to form N-retinylidene-PE. The acidic pH of the disc lumen can trap the protonated form of N-retinylidiene-PE in the lumen leaflet of the disc membranes [111
]. The function of ABCA4 is envisioned to actively transport or flip trapped N-retinylidene-PE from the lumen to the cytoplasmic side of the disc membrane. On the cytoplasmic surface, N-retinylidene-PE can dissociate into all-trans
retinal and PE, and all-trans
retinal can then be reduced to all-trans
retinol by RDH8 for the resynthesis of 11-cis
retinal and the regeneration of rhodopsin via the visual cycle. In this model the function of ABCA4 is to insure that all-trans retinal and N-retinylidene-PE is completely removed from disc membranes following photobleaching of rhodopsin in rod cells and cone opsin in cone cells.
Loss or decrease in ABCA4 transport activity will result in an accumulation of N-retinylidene-PE and all-trans
retinal in disc membranes as found in abca4
knockout mice and Stargardt patients [70
] (, ). N-retinylidene-PE can react with all-trans
retinal to form the diretinal derivative A2PE () and related diretinal compounds in outer segments [71
]. Rod and cone outer segments undergo a continual renewal process in which new membrane is added at the base of the outer segments while packets of aged outer segments are removed by shedding and phagocytosis by adjacent RPE cells. Through this process, the outer segment is renewed once every 10 days. Ingested phagosomes containing A2PE, all-trans retinal dimers and other retinoid compounds fuse with lysosomes to form phagolysosomes which contain enzymes that degrade the outer segments (). A2PE is hydrolyzed to A2E, but cannot be further metabolized. As a result, A2E and related retinal compounds progressively accumulate as lipofuscin deposits in RPE cells [71
]. Upon exposure to blue light, A2E can be converted into epoxides [79
]. A2E and related epoxides can lead to loss in RPE function and cell death [75
]. Since RPE cells are required for photoreceptor cell survival, the loss in RPE cells will cause the photoreceptor cell death and a loss in vision.
This model is consistent with many of the characteristic features observed for individuals with Stargardt disease and mice deficient in ABCA4. These include elevated levels of all-trans retinal and protonated N-retinylidiene-PE in photoreceptor cells, and the progressive accumulation of A2E and other diretinal compounds as fluorescent lipofuscin deposits in RPE cells. In addition the accumulation of all-trans retinal in disc membranes can explain the delay in dark adaptation found in Stargardt patients and abca4 mice.
Although there is considerable support for this model in which ABCA4 functions as N-retinylidene-PE transporter, a number of issues remain to be resolved. In particular, it is necessary to directly show that ABCA4 functions to transport or flip N-retinylidene-PE across disc membranes. The mechanism by which ATP hydrolysis is coupled to the transport of N-retinylidene-PE across membranes remains to be elucidated. Importantly, the direction of substrate transport needs to be clarified. In the present model, ABCA4 is envisioned to transport N-retinylidene-PE from the lumen to the cytoplasmic leaflet of disc membranes. This is the opposite direction of transport for many well studied mammalian ABC transporters. ABC transporters are known to be regulated through protein-protein interactions and/or phosphorylation. To date there is no information on factors which regulate the activity of ABCA4. Resolution of these issues awaits the further development of functional transport assays and the generation of high resolution structures of ABCA4.