Distribution Patterns of Photoreceptors, Protein, and Cyclic Nucleotides in the Human Retina
Total Page:16
File Type:pdf, Size:1020Kb
Distribution Patterns of Photoreceptors, Protein, and Cyclic Nucleotides in the Human Retina Debora B. Farber, fj John G. Flannery,t$ Richard N. Lolley,*§ ond Dean Dok*tt The concentration of cGMP, cAMP, protein and the number of cone and rod photoreceptors have been measured in parallel arrays of punches, 3 mm in diameter, taken from each quadrant of normal human retinas. A separate punch containing the fovea and parafoveal region was also analyzed. Eyes were obtained from four male donors ranging in age from 35 to 67 yr. The retina thins considerably from the center to the periphery, and consequently the protein content forms a gradient in the same direction. Similar gradients were observed for cAMP and cGMP concentrations. In all eyes studied, the foveal- parafoveal region had higher levels of cAMP than cGMP. The data was analyzed with the aid of a computer in order to obtain three-dimensional maps of the patterns of distribution of the different parameters. A strong correlation between the areas of higher cone density, non-photoreceptor neurons, and cAMP, and an equally strong correlation between rod distribution and that of cGMP was observed. These maps will serve as baseline data in studies of pathological conditions such as retinitis pigmentosa. Invest Ophthalmol Vis Sci 26:1558-1568, 1985 A large proportion of human blindness is due to within specific regions. This regional vulnerability of inherited diseases of the retina. Studies directed toward photoreceptors necessitates the biochemical sampling elucidating the causes of these disorders have been re- of numerous sites within the fundus if specific abnor- stricted by limited availability of human tissue at a malities are to be correlated with the pathological area. time when it could render useful information concern- In a previous attempt by others at achieving this goal,4 ing possible metabolic abnormalities. It has been shown tissue sampling involved the collection of three con- that disordered metabolism of the cyclic nucleotides, centric rings of retinal tissue, using the foveola as the cyclic GMP (cGMP) and cyclic AMP (cAMP), occurs geometric center of a circle. In this way, the posterior, in some animals with inherited retinal degenerations.1"3 central, and peripheral regions of the retina were an- However, it is unknown whether abnormalities in cyclic alyzed independently. In the current study, we sought nucleotides are involved in any forms of human blind- to improve the resolution of retinal sampling. Data ness. This uncertainty can be resolved only if human were collected from parallel linear arrays of punches, retinas from early stages of disease are analyzed bio- 3 mm in diameter, taken from each quadrant. The chemically in a systematic manner with foreknowledge data obtained in this way were analyzed with the aid of the normal distribution patterns of cyclic nucleo- of a computer in order to demonstrate relationships tides. within the fundus. We present here numerical data as Our interest in retinitis pigmentosa has stimulated well as graphic reconstructions of the distribution pat- the current study in which microtechniques were de- terns of protein, cGMP, cAMP and rod and cone pho- veloped for the morphological and biochemical analysis toreceptors. The biochemical observations deal with of loci within normal or diseased human retinas. In- the pattern of protein, cyclic AMP and cyclic GMP herited human retinal degenerations usually arise distribution throughout the neural retina, and the morphological data complement and expand upon the classic work of Osterberg.5 The baseline data generated in this study will be used for comparison with results From the Departments of Anatomy* and Ophthalmology! and Jules Stein Eye Institute^ University of California, Los Angeles School obtained from donor eyes affected with retinitis pig- of Medicine, Los Angeles, California; and Laboratory of Develop- mentosa whenever they become available. mental Neurology,§ Veterans Administration Medical Center Se- pulveda, California. Supported by USPHS Grant EY-00444, EY-00331, EY-02651, Materials and Methods EY-05625, EY-00395 and a Center Grant from the National Retinitis All of the eyes used for this study were deemed nor- Pigmentosa Foundation, Fighting Blindness, Inc., Baltimore. mal prior to surgery. They were obtained from patients Submitted for publication: April 2, 1985. Reprint requests: Dr. John G. Flannery, Jules Stein Eye Institute, who had cancer of the eyelid that had extended pos- UCLA School of Medicine, Los Angeles, CA 90024. teriorly, and in whom orbital exenteration was deemed 1558 Downloaded from iovs.arvojournals.org on 10/02/2021 No. 11 PHOTORECEPTOR5 AND CYCLIC NUCLEOTIDES IN HUMAN RETINA / Forber er d. 1559 necessary to save the life of the patient. The patients had received no anticancer drugs prior to surgery and ranged in age from 35 to 67 yr. Eyes from four male patients were used in this study. After removal of the cornea, lens and iris, incisions were made in order to divide the eyecup into four separate quadrants: superior and inferior temporal, and superior and inferior nasal (Fig. 1). The incisions were carried close to, but did not include, the optic nerve head, so that all four quad- rants remained attached to one another at the posterior pole of the eye. With the aid of a 3-mm trephine (Storz; St. Louis, MO), a sample that included the foveola, fovea and parafovea was taken from the central neural retina. Then, starting from the peri fovea and proceed- ing to the peripheral retina, four sequential and parallel 3-mm punches of the neural retina were taken from the superior temporal quadrant. Equivalent parallel samples were taken from the other three quadrants as Fig. 1. Diagram illustrating the pattern in which tissue was collected well. Each of the samples was analyzed for cAMP, from donor eyes for biochemical analysis. The eyecup is depicted following removal of the anterior segment and separation into superior cGMP and protein. The remainder of the eyecup was temporal (ST), inferior temporal (IT), superior nasal (SN) and inferior fixed overnight in 1% formaldehyde and 2% glutaral- nasal (IN) quadrants. Neural retina containing foveola, fovea and dehyde in O.I M sodium phosphate buffer, pH 7.3. parafovea was taken from the most central portion of the ST quadrant with a 3 mm trephine. Thereafter, four parallel punches of neural Single, 3-mm punches of fixed neural retina parallel retina were removed successively from each quadrant in a central to to the dual samples made for biochemistry were taken peripheral direction up to the ora serrata. After samples were collected from each quadrant for photoreceptor counts. All for biochemistry, the retina was fixed in aldehydes, and additional punches were fixed additionally with 1% osmium te- samples were collected for morphological analysis. The optic nerve troxide in 0.1 M sodium phosphate buffer for one hour, head is indicated by the cross-hatched circle in the center of the figure. The diagram is schematic and not drawn to scale. Tissue sam- dehydrated in graded ethanol (50-100%), cleared in ples were more closely-spaced than indicated. propylene oxide and embedded in Araldite 502 (Ciba Products Co.; Summit, NJ). Cross sections of each punch were cut through the photoreceptor inner or central punch taken from the 71-year-old donor. The outer segments. Because there is only one fovea per data from these counts is summarized in Table 1. retina, parallel biochemical and morphologic samples Three-dimensional plots of photoreceptor distribution could not be taken for the central sample. We utilized were constructed from these data and are primarily the central punch for biochemical analysis exclusively, since the primary emphasis of this study was to establish 2 baseline biochemical measurements. The central retinal Table 1. Mean number of photoreceptors per mm photoreceptor counts reported in this study were ob- (± one standard error of the mean) tained from an additional aldehyde-fixed donor eye Inferior nasal Inferior temporal representing the upper age limit (67 yr) of the other specimens. The foveola of a 71-year-old donor eye was in 1: rods: 135025 ±10082 ill: rods: 168923 ±8604 located in radial sections and then cross-sectioned in cones: 6358 ± 382 cones: 7563 ± 755 order to make photoreceptor counts. One micron sec- in2:rods: 133269 ± 17500 it2:rods: 159192 ± 1 1654 cones: 5538 ±1154 cones: 5730 ± 269 tions from all samples were placed on microscope in3:rods: 111692 ± 15813 it3:rods: 91641 ±36017 slides, de-plasticized with 1/6 saturated sodium ethox- cones: 6127 ± 739 cones: 5255 ± 622 ide, and stained with Toluidine Blue. Photomicro- in4: rods: 73666 ± 13975 it4: rods: 76743 ± 9466 cones: 5538 ± 789 graphs of each section were taken with a Zeiss Pho- cones: 3666 ± 321 tomicroscope III (Carl Zeiss; Oberkochen, West Ger- Superior nasal Superior temporal many) at a magnification of 252X, and the negatives snl:rods: 156876 ± 17739 stl:rods: 145718 ±9135 were enlarged 6.5 X. Ten equal areas, randomly selected cones: 6307 ± 400 cones: 6563 ± 606 from each punch and showing the best cross-sectional sn2: rods: 141435 ±23883 st2:rods: 127128 ±23029 cell orientation, were selected, and the rods and cones cones: 6076 ± 1050 cones: 7494 ± 668 sn3:rods: 123692 ± 13539 st3: rods: 137384 ± 9459 counted. The means and standard errors of the pho- cones: 5743 ± 672 cones: 4897 ± 200 toreceptor counts were then estimated for each of the sn4: rods: 84026 ± 19861 st4:rods: 85128 ±27930 16 areas evaluated in each of the eyes, including the cones: 4768 ± 1250 cones: 3743 ±916 Downloaded from iovs.arvojournals.org on 10/02/2021 1560 INVESTIGATIVE OPHTHALMOLOGY & VI5UAL SCIENCE / November 1985 Vol. 26 Fig. 2A and B. Photomicrographs prepared from cross sections of the rod-cone mosaic in the pericentral (non-foveal) and peripheral {near ora serrata) retina of the 49-year-old patient used in this study.