A POSSIBLE ULTRASTRUCTURAL CORRELATE OF FUNCTION IN THE FROG RETINAL ROD.
J. Robertson
DOI: https://doi.org/10.1073/pnas.53.4.860
Abstract:Previous studies ofvertebrate retinal receptors have established that each receptor lamnella consists of two unit membranes in close apposition along their outside surfaces. 1' 2 It was shown that each cone lamella is a simple invagination of thesur face membrane of the outer segment,1' 37 and that continuity is maintained in the adult retina over some extent of the circumference of each lamella. This is true of cones of frogs,",7 toads,5'6 perch,3Imonkeys,8 and pigeons.9 A similar type of arrangement was described in frog rod outer segments, but only at the base of the rod, near the inner segment.1 8-9 Here, the first few lamellae were found to be simple ilivaginatioins of the surface membrane of the outer segment. The remaining lamnellae seemed to consist of unit mnemnbranses united ill a five-layered compound membrane structure about 150 A thick, but Iio evidence of direct conltiniuity between the lamellae, either with one another or via the surface membranes of the outer segments, was observed. Several other investigations of rods in frogs10-'2 and other animals'l3-6 have also failed to yield evidence of continuity either between unit membranes of adjacent lamiiellae or between the lamiellae and the surface membranes. To be sure, it has been postulated fromn time to time that there might be some type of continuity between laimiellae, taking the form, perhaps, of narrow, tubular coinnections runninlg front one lamella to the adjacent one, as claimed by Sjdstrand for perch rods." However, his micrographs are miost unllconvincinlg and we agree with Cohen'8 that a clear demonstration of such microvillous interconnections between lamellae has never been made. The concept of each rod lamella's being a closed, disc-like flattened sac has gained wide acceptance (see ref. 18). In considering mechanisms whereby the absorption of one photon might lead to a propagated action potential in the rod cell, there are certain restrictions imposed by the discontinuous isolated disc nature of the rod laiimellae. These have been discussed recently by Brown, Gibbons, and Wald."5 Such mechaiiisms as exciton migration' and photoconductions were considered reasonable for the initial spread of excitation, but such mechanisms were objected to on the grounds that they would probably be confined to a single isolated lamiiella and fail to excite the whole structure. Such considerations have led to the notion that the plasma membrane of the outer segment is the probable site of initial neural excitation, even though it appears probable, for a number of reasons, that the initial photon absorption occurs ill the lanmellae. The results reported here suggest that the lamellae of rods are not isolated discs, but rather parts of one continuous membrane that is, in turn, conltinuous with the plasma mnemiibrane at the base of the rod. Hence, such mechanisms as excitonI migration and photoconductioni become more plausible. The results also suggest that the rod unit membranes are physical extensions of the surface inembralie of the rod cell. Hence, the possibility of studying highly oriented surface membranes directly