Mechanism for the selective uptake of macular carotenoids mediated by the HDL cholesterol receptor SR-BI.

The macular carotenoids lutein and zeaxanthin are taken up from the bloodstream into the human retina through a selective process, for which the HDL cholesterol receptor scavenger receptor BI (SR-BI) in the cells of retinal pigment epithelium (RPE) is thought to be a key mediator. However, the mechanism of SR-BI-mediated selective uptake of macular carotenoids is still not fully understood. Here, we investigate possible mechanisms using biological assays and cultured HEK293 cells, a cell line without endogenous SR-BI expression. Binding affinities between SR-BI and various carotenoids were measured by surface plasmon resonance (SPR) spectroscopy, which shows that SR-BI cannot bind lutein or zeaxanthin specifically. Overexpression of SR-BI in HEK293 cells results in more lutein and zeaxanthin taken up than ß-carotene, and this effect can be eliminated by an SR-BI mutant (C384Y) whose cholesterol uptake tunnel is blocked. Next, we determined the effects of HDL and hepatic lipase (LIPC), SR-BI's partners in HDL cholesterol transport, on SR-BI-mediated carotenoid uptake. HDL addition dramatically reduced lutein, zeaxanthin, and ß-carotene in HEK293 cells expressing SR-BI, but the cellular lutein and zeaxanthin are higher than ß-carotene. LIPC addition increases the uptake of all three carotenoids in HDL-treated cells, and promotes the transport of lutein and zeaxanthin better than ß-carotene. Our results suggest that SR-BI and its HDL cholesterol partner HDL and LIPC may be involved in the selective uptake of macular carotenoids.

Imaging lutein and zeaxanthin in the human retina with confocal resonance Raman microscopy.

Lutein and zeaxanthin are xanthophyll carotenoids that are highly concentrated in the human macula, where they protect the eye from oxidative damage and improve visual performance. Distinguishing lutein from zeaxanthin in images of the human retina in vivo or in donor eye tissues has been challenging because no available technology has been able to reliably differentiate between these two carotenoids, which differ only in the position of one C = C bond. Here, we report the differential distributions of lutein and zeaxanthin in human donor retinas mapped with confocal resonance Raman microscopy. Zeaxanthin is highly concentrated in the fovea, extending from the inner to the outer limiting membranes, with especially high concentrations in the outer plexiform layer, while lutein is much more diffuse at relatively lower concentration. Our results imply that zeaxanthin may play a more important role than lutein in human macular health and disease.

Supplementation with macular carotenoids improves visual performance of transgenic mice.

Carotenoid supplementation can improve human visual performance, but there is still no validated rodent model to test their effects on visual function in laboratory animals. We recently showed that mice deficient in ß-carotene oxygenase 2 (BCO2) and/or ß-carotene oxygenase 1 (BCO1) enzymes can accumulate carotenoids in their retinas, allowing us to investigate the effects of carotenoids on the visual performance of mice. Using OptoMotry, a device to measure visual function in rodents, we examined the effect of zeaxanthin, lutein, and ß-carotene on visual performance of various BCO knockout mice. We then transgenically expressed the human zeaxanthin-binding protein GSTP1 (hGSTP1) in the rods of bco2-/- mice to examine if delivering more zeaxanthin to retina will improve their visual function further. The visual performance of bco2-/- mice fed with zeaxanthin or lutein was significantly improved relative to control mice fed with placebo beadlets. ß-Carotene had no significant effect in bco2-/- mice but modestly improved cone visual function of bco1-/- mice. Expression of hGSTP1 in the rods of bco2-/-mice resulted in a 40% increase of retinal zeaxanthin and further improvement of visual performance. This work demonstrates that these "macular pigment mice" may serve as animal models to study carotenoid function in the retina.

Structure of the lutein-binding domain of human StARD3 at 1.74†Šresolution and model of a complex with lutein.

A crystal structure of the lutein-binding domain of human StARD3 (StAR-related lipid-transfer protein 3; also known as MLN64) has been refined to 1.74 Šresolution. A previous structure of the same protein determined to 2.2 Šresolution highlighted homology with StARD1 and shared cholesterol-binding character. StARD3 has since been recognized as a carotenoid-binding protein in the primate retina, where its biochemical function of binding lutein with specificity appears to be well suited to recruit this photoprotective molecule. The current and previous structures correspond closely to each other (r.m.s.d. of 0.25 Å), especially in terms of the helix-grip fold constructed around a solvent-filled cavity. Regions of interest were defined with alternate conformations in the current higher-resolution structure, including Arg351 found within the cavity and Ω1, a loop of four residues found just outside the cavity entrance. Models of the complex with lutein generated by rigid-body docking indicate that one of the ionone rings must protrude outside the cavity, and this insight has implications for molecular interactions with transport proteins and enzymes that act on lutein. Interestingly, models with the ℇ-ionone ring characteristic of lutein pointing towards the bottom of the cavity were associated with fewer steric clashes, suggesting that steric complementarity and ligand asymmetry may play a role in discriminating lutein from the other ocular carotenoids zeaxanthin and meso-zeaxanthin, which only have ß-ionone rings.