In vivo observation of human retinal-cone outer-segment renewal

Vision is one of the most complex human senses. The eye projects images from our environment onto the retina, which consists of a variety of different cells. The retina preprocesses this information before sending it to the brain as electrical signals. Many different retinal diseases—such as age-related macular degeneration—distort the fragile retinal structure, leading to (for example) reduced visual perception and eventual blindness.

Maintenance of a healthy retina requires interactions between its photosensitive cells and the retinal-pigment epithelium (RPE). There are two different types of retinal photosensitive cells, including cones (for photopic vision under well-lit conditions) and rods (for scotopic vision under poor lighting). Each consists of four parts, i.e., inner segments (ISs), outer segments (OSs), a connecting cilium (IS/OS), and a synaptic terminal for neuronal signal propagation. Photo pigments are contained within stacks of membrane disks within the OSs of cone cells. To minimize damage by photo-toxic processes, the OSs change with time,¹ which includes formation of new disks at the IS/OS, disk shedding, and subsequent phagocytosis of shed disks at the border between photoreceptors and the RPE. Species- and photoreceptor-dependent differences in these processes have been reported from extensive in vitro studies in animals.¹ Observing these processes in vivo is a challenge, since it requires both 3D cellular resolution and imaging of exactly the same location of the retina over extended periods of time. Additionally, at high magnification, eye motion is an inherent problem.

Figure 1 shows a typical depth-integrated optical coherence tomography (OCT) image of the cone mosaic and a representative OCT cross-sectional (B-) scan illustrating different parts of the cone photoreceptors.² We used a transverse-scanning OCT/scanning-laser ophthalmoscopy instrument to visualize the retina at high resolution over extended periods of time. Together with an axial eye tracker,³ the high acquisition speed (40 frames/s) of the instrument allowed us to record 3D volumes of the retina with few eye-motion artifacts.⁴ Interestingly, we observed bright reflection spots (BRSs) of unknown origin within a few percent of the OSs of cone photoreceptors.⁴

Figure 1. (a) Depth-integrated optical coherence tomography (OCT) image of the cone mosaic, at ∼4°eccentricity from the fovea. White spots correspond to single cones. Image extension: ∼300×300μm². (b) Representative OCT cross-sectional scan showing different parts of cone photoreceptors. IS/OS: Connecting cilium junction between inner and outer photoreceptor segments. ETPR: End tips of photoreceptors. RPE: Retinal-pigment epithelium (hardly visible because of linear gray scale).

To shed some light on their origin, we recorded 3D data sets of the retinas of healthy volunteers for different durations up to several days. We registered these data sets to each other with better accuracy than the extension of a single cone. We recorded two different measurement series. In the first, 3D volumes were recorded every hour for seven hours. In the second, 3D volumes were recorded once each day for three days. In the first series, we only observed intensity changes within the OCT B-scans, while the location of the IS/OS and end tips of photoreceptors (ETPR) remained unchanged. However, in the second series—in addition to intensity fluctuations—we observed changes in the ETPR position. We believe that this corresponds to different phases of the renewal process for each individual cone. New OS disks are generated at the IS/OS junction, probably giving rise to the observed intensity fluctuations. Occasionally, the RPE sheds disks at the distal end of the cones (the position of the ETPR), which are then disintegrated by phagocytosis within the RPE. These two processes differ considerably in their dynamics: generation of new disks is a continuous process, while phagocytosis is not. Therefore, depending on the time delay between the last disk-shedding event and measurement of each individual cone, different OS lengths can be observed, because the number of newly generated disks varies.

We observed movement of several BRSs toward the ETPR within the recorded data sets (see Figure 2).² We hypothesize that these BRSs originate from cracks or defects within the packing structure of the OS disks, resulting in changes in refractive index and giving rise to the observed OCT signal. The crack moves toward the ETPR because of generation of disks at the IS/OS boundary and disk shedding by the RPE (i.e., the cone-renewal process). Finally, we measured the average speed of all BRS motions in two volunteers and found an OS growth rate of 110±40nm per hour, which is in excellent agreement with the renewal rate measured in animal studies^5,6 and results indirectly obtained using an adaptive-optics-equipped fundus camera.⁷

Figure 2. OCT cross-sectional scans reveal changes of cone photoreceptors over 96h. A single bright reflection spot (BRS) appears at 48h and moves toward the ETPR at 72 and 96h.

Based on these studies, changes in outer-segment renewal in different diseases may be observable, leading to better understanding of retinal-disease development. Future studies are needed to measure individual fluctuations of cone renewal. This represents part of our ongoing efforts.