Photoreceptor shedding can be initiated within the eye

Abstract

IN the vertebrate retina the light-sensitive rod outer segments (ROS) are composed of stacks of double membrane ‘disks’ surrounded by a plasma membrane. Throughout life, the ROS membranes are renewed in a highly ordered manner which has made this cell type the model of choice for many studies on membrane synthesis and turnover. Autoradiographical studies have shown that new membrane precursors are synthesised within the rod inner segment, then assembled into new disks at the base of the ROS, displacing previously formed disks away from the cell body^1,2. An optimum ROS length is maintained by the periodic shedding of groups of older disks from the ROS tip. These discarded disk packets are then engulfed and degraded by the adjacent pigment epithelium^3,4. The shed ROS fragments persist for several hours within the cytoplasm of the pigment epithelial cells and are referred to as phagosomes.³ LaVail⁵ has reported that ROS in the rat retina shed their apical tips shortly after the onset of light in animals maintained under cyclic lighting (12h light (L):12h dark (D)). We have shown that a similar pattern of rod outer segment shedding occurs in the amphibian retina^6–8, and similar patterns have also been observed in the goldfish⁹, rabbit¹⁰ and chick¹¹. In all these animals, a nearly synchronous burst of ROS shedding occurs within two hours of the onset of light, and minimal amounts of rod shedding occur during the remainder of the day. Also, in amphibians, constant light inhibits daily rod shedding, and the outer segments elongate due to continued outer segment disk addition^7,12,13. In the rat⁵ and Xenopus⁷, shedding occurs at approximately the same time each day in animals maintained for several days in the dark, although in Rana the onset of light is required to initiate shedding, and dark-maintained animals do not shed for several days⁶. Taken together, these results have suggested that ROS shedding follows a circadian rhythm related to the onset of the light cycle. The photoreceptive nature of the amphibian pineal gland, and the well-documented circadian rhythm in serotonin–melatonin metabolism in the rat pineal, made this organ a particularly good candidate for controlling photoreceptor shedding, although several other extraocular systems have been suggested¹⁴. However, several laboratories have demonstrated that pinealectomy, hypophysectomy, superior cervical ganglionectomy, and cutting of the optic nerves have no effect on the diurnal pattern of ROS shedding in either the rat or the frog retina^12,15. Furthermore, we have been unable to affect rod shedding in the frog retina by direct injection of melatonin, serotonin or the β-adrenergic effectors propranolol and iso-prenaline. Thus, even though rod outer segment shedding consistently occurs in both eyes and seems to follow a circadian rhythm in darkness, suggesting some form of central regulation, the intraocular control of this response remained a possibility. We report here experiments which strongly suggest that rod shedding in frogs can be initiated within the eye and is not secondarily controlled by a gland or organ elsewhere in the body.