* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Vesicular transport of newly synthesized opsin from the Golgi
Survey
Document related concepts
Node of Ranvier wikipedia , lookup
Cell nucleus wikipedia , lookup
Cell encapsulation wikipedia , lookup
Magnesium transporter wikipedia , lookup
Lipid bilayer wikipedia , lookup
Membrane potential wikipedia , lookup
Chemical synapse wikipedia , lookup
Organ-on-a-chip wikipedia , lookup
Signal transduction wikipedia , lookup
Model lipid bilayer wikipedia , lookup
Cytokinesis wikipedia , lookup
SNARE (protein) wikipedia , lookup
List of types of proteins wikipedia , lookup
Transcript
Vesicular Transport of Newly Synthesized Opsin from the Golgi Apparatus toward the Rod Outer Segment Ulrrostrucrurol Immunocyrochemicol and Auroradiographic Evidence in Xenopus Retinas David 5. Papermasrer, Barbara G. Schneider, and Joseph C. Desharse* Each day, rod photoreceptors of the vertebrate retina synthesize rhodopsin and insert it into new membranes of the rod outer segment (ROS). The authors determined which components of the rod cell transport opsin from the Golgi to the ROS by a combined EM autoradiographic and immunocytochemical study using radiolabeled amino acid precursors and antiopsin antibodies. Radiolabeled proteins in the ellipsoid region of Xenopus laevis retinal rods were localized by comparison of the distribution of silver grains with the predicted distribution generated by a hypothetical source: grain matrix. Sources of decay were not uniformly distributed. Small vesicles compressed between mitochondria and clustered beneath the connecting cilium that joins the inner to the outer segment contained more than 30% of the radiolabel and had a specific activity 17 times higher than the surrounding cytoplasm. Opsin was localized immunocytochemically on thin sections of retinas embedded in Lowicryl K4M (Polysciences; Warrington, PA) by reaction sequentially with biotinyl-rabbit antifrog opsin, biotinyl-sheep antirabbit F(ab')2, and avidin-ferritin. Golgi apparatus, intermitochondrial vesicles, and vesicles that clustered beneath the connecting cilium were prominently labeled. Subellipsoid smooth endoplasmic reticulum was labeled at background levels. These results demonstrate that intracellular vesicular membranes transport newly synthesized opsin from the Golgi to the base of the connecting cilium of X. laevis retinas. Antibody labeled the outer segment plasma membrane at a 10-fold greater density than the contiguous inner segment plasma membrane. The polarized distribution of opsin apparently involves not only vectorial transport of opsin in the inner segment but also restrictions to the randomization of opsin inserted into the inner and outer segment plasma membrane. Invest Ophthalmol Vis Sci 26:1386-1404, 1985 cell surfaces.1 4 The polarized budding of virus-infected cells probably involves vesicular transport of virus membrane proteins to the cell surface.5"13 Similar pathways may function in the transport of newly synthesized membrane proteins in photoreceptor cells. Rod photoreceptor cells offer, however, several advantages for the study of the biosynthesis, processing, and sorting of membrane proteins. The major protein synthesized in the entire retina is the visual pigment apoprotein, opsin. Its synthesis is a normal physiologic function of the rod cell. Rod photoreceptors assemble an extraordinary amount of new outer segment disk membranes. Under normal circadian light cycles, about 80 disks are formed per day in Xenopus laevis tadpole rods. 1415 Each disk is about 6 nm in diameter or nearly the size of a human red cell and contains 106 rhodopsins/disk. This corresponds to the generation of about 4500 /um2 of new The generation and maintenance of cell polarity in epithelial and neuronal tissues suggests that specific cell membrane constituents are uniquely transported or are otherwise restricted to sites of function in these cells. In epithelial cells, vesicles may participate in the transport of hormones, immunoglobulins, and serum proteins across the cell and from sites of synthesis to unique From the Department of Pathology, VA Medical Center, West Haven, and Yale Medical School, New Haven, Connecticut; and the Department of Anatomy and Cell Biology, *Emory University School of Medicine, Atlanta, Georgia. Supported in part from NIH grants EY-03239, EY-00845, GM21714, EY-02414, EY-03222, and the Veterans Administration. During portions of this research DP and JB were recipients of RCDA grants EY-00017 and EY-00169, respectively. Submitted for publication: October 11, 1984. Reprint requests: David S. Papermaster, MD, Department of Pathology/113, VA Medical Center, West Haven, CT 06516. 1386 Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. 1387 F. Disk Assembly Fig. 1. Diagram of X. laevis rod photoreceptor inner and outer segment illustrating the cellular membranes involved in the biosynthesis and transport of opsin. PostGolgi vesicles vectorially transport newly synthesized opsin past the closely packed mitochondria in the ellipsoid region. Vesicles apparently fuse with the grooves of the periciliary ridge complex near the base of the connecting cilium. Opsin may then proceed along the ciliary plasma membrane to become incorporated into the new disks forming at the base of the outer segment. ROS E. Ciliary Transport D. Insertion at Plasma Membrane of the PRC Basal Bodies C. Vesicular Transport Ellipsoid _ SER B. GolghTerminol Processing, Glycosylation and Packaging photoreceptor membrane daily.15 It represents an average rate of 3.1 /xm2/min of rod outer segment (ROS) membrane addition, a rate which is comparable to that occurring at the growth cones of actively elongating neurites.16 Rod outer segments contain no cellular constituents for protein synthesis. Synthesis of disk membrane proteins occurs in the inner segment. Since rhodopsin is a typical intrinsic membrane protein within ROS disks, we wished to determine which cell constituents might participate in its transport across the inner segment. Prior autoradiographic and radiobiochemical studies showed that most of the radiolabeled protein migrated from the Golgi past the ellipsoid—a mitochondria rich domain between the nucleus and the ROS—to arrive eventually in newly formed basal.ROS disks.17"19 Photoreceptor inner and outer segments are joined by a narrow connecting cilium. No membranes were seen in the interior of the cilium, yet radiolabeled proteins were shown to congregate beneath its base and silver grains were present over cytoplasm beneath the cilium and within the cilium interior. This was interpreted as possible evidence for a soluble form of opsin as an intermediate during its transport and final passage through the cilium.20 Subcellular fractionation of frog retinas demonstrated, however, that newly synthesized opsin was isolated only on easily sedimented membrane fractions. The cytosol fractions of retinal homogenates did not R E R : Synthesis and Core Glycosylation Myoid contain opsin.21 Vesicles were postulated as possible carriers of newly synthesized opsin from the Golgi to the base of the connecting cilium.22 In amphibian retinas, vesicles and cisternae are observed in the inner segment, especially in the ellipsoid and in the cytoplasm adjacent to the basal body of the connecting cilium.23"25 Additional evidence favoring a precursor role for the periciliary vesicles was obtained by freeze fracture analysis of their structure. Besharse and Pfenninger23 showed that intramembranous particles in the periciliary vesicles had 10-nm diameters that were comparable to particles in outer segment disks. Their diameters were distinct from those in adjacent mitochondria and inner segment plasma membranes. To evaluate further the possible function of these vesicles, we began a joint effort of electron microscopic autoradiographic and immunocytochemical analysis of thin sections of X. laevis juvenile, tadpole, and adult retinas. Our results indicate that radiolabeled protein is closely associated with ellipsoid and periciliary vesicle membranes after 2 hr of incorporation. Antibodies to opsin bind to these post-Golgi vesicles in the ellipsoid. Together, these results indicate that some of these vesicles contain newly synthesized opsin destined for the ROS.* These observations are introduced in Figure 1. * Portions of these results were presented in preliminary form at a meeting of the American Society for Cell Biology (Papermaster DS, Schneider BG, and Besharse JC: J Cell Biol 83:275a, 1979). Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1388 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / October 1985 Materials and Methods Xenopus laevis tadpoles, juveniles, and adults were maintained at 20-25 °C under a 12 hr light: 12 hr dark cycle. Some tadpoles were maintained for 6 days in darkness and exposed to light for 30 to 120 min. Tadpoles at larval stages 54-56 26 were injected with a single dose of 50 ;uCi of L-[(N)-4,5-3H] leucine (New England Nuclear; Boston, MA) with a specific activity of 54.6 Ci/mmol. Postmetamorphic juveniles (3.5 cm long) received a mixture of tritiated amino acids consisting of 100 j*Ci of L-[(N)-4,5-3H] leucine (specific activity, 50 Ci/mmol) from ICN Chemical and Radioisotope Division, Irvine, California, and 50 ixd each of L[2,3,4-3H] valine (specific activity, 11.1 Ci/mmol) and L-[ring-2,3,4,5,6-3H] phenylalanine (specific activity, 98.6 Ci/mmol) from New England Nuclear. Radioactive amino acids were delivered by intraperitoneal injections using 30-gauge disposable needles at 15 to 60 min prior to light onset under red light (Wratten No. 2 filter, Kodak; Rochester, NY). Fixations were in darkness just before light onset or at 30, 60, and 120 min after light onset. Eye cups prepared by surgical removal of cornea, iris, and lens were fixed either in 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer at pH 7.4 (tadpoles), the same aldehydes in 0.067 M cacodylate buffer (postmetamorphic juveniles), or 4% paraformaldehyde in 0.1 M cacodylate buffer (tadpoles). The last fixation was carried out to control for possible nonspecific binding of radiolabeled amino acids.27 Portions of the eye were embedded in Spurr's medium (postmetamorphic juveniles)28 or Epon-Araldite (tadpoles) for autoradiography,15 and in bovine serum albumin (BSA)29-30 or Lowicryl K4M (Polysciences; Warrington, PA) 3132 for immunocytochemistry. These investigations were carried out in accordance with the ARVO Resolution on the Use of Animals in Research. EM Autoradiography Thin sections (silver-gold interference colors) of retina were obtained; they were assumed to be 100 nm thick for purposes of quantitative analysis. Sections were coated on glass slides with a monolayer of Ilford L-4 emulsion using the flat substrate method.33 Following exposure in black dessicator boxes, sections were developed with Kodak D19 or Microdol X, Phenidonascorbic acid, or with Agfa-Gevaert fine grain developer after gold latensification.34 Although the fine grain developers provided an excellent qualitative understanding of the distribution of radioactivity, the detailed quantitative analysis presented in the results was con- Vol. 26 ducted on sections developed with Microdol X because such material has been thoroughly characterized with regard to resolution and use in quantitative studies.35 We used a resolution value in units of half-distance (HD = 157 nm) appropriate for our section thickness.35 Comparable analyses (not reported in results) were carried out using tissues fixed in paraformaldehyde only and on paraformaldehyde-glutaraldehyde fixed tissue developed with Kodak D19 or Phenidon-ascorbic acid. Although each of the three analyses involved fewer observed silver grains than the results described in Table 1, the results were consistent with those described in Table 1 for Microdol X development of tissues fixed with mixed aldehydes. Thin sections passing through a plane approximately parallel to the long axis of photoreceptors were used to obtain a collection of micrographs for quantitative analysis. The inner segment and basal outer segment region of all cells visible in sections was photographed and printed at a final magnification of XI 8,000. The use of all visible cells rather than a selected midcellular sample yielded a random collection of cells sectioned in different axial planes. In order to estimate the relative degree of labeling of various sources within photoreceptor inner segments, we used the hypothetical grain distribution method of Blackett and Parry36'37 as developed further by Salpeter et al.38 In this method a hypothetical sourcergrain matrix was generated using masks designed to take into account the extent to which radioactivity in a structure would be expected to contribute silver grains over all adjacent structures. This matrix was generated by identifying possible real sources of radioactivity (rows in Table 1). Silver grains resulting from hypothetical disintegrations were assigned to grain compartments (columns in Table 1). Observed silver grains were then tabulated using the same definition of grain compartments as was used for generation of the hypothetical source-grain matrix; the sources corresponding to real grains were, of course, unknown. Although direct comparison of the hypothetical distribution with the real distribution was useful and was formally analogous to comparisons made using the probability circle method of Williams,39 the real value of this approach was to obtain an estimate of the source density of radioactivity within the individual organelles identified as sources (Tables 2 and 3). The source densities were estimated by using a computer program which systematically varied a series of multipliers (source density values) that altered the hypothetical source-grain matrix until the x 2 value in comparing real and expected distributions was minimized. The final parameters were then considered to be source Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1389 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. No. 10 Table 1. Localization of 3H-labeled protein in the ellipsoid of rod inner segments (source: grain matrix) Grain compartments PM Cytoplasm Mitochondria Mitochondriacytoplasm Vesiclecytoplasm 10 258 6.1 18 19 3.0 140 33 20.0 114 84 3.8 33 3 2.2 327 44 1.9 46 10 4.7 214 2006 7.8 29 8 6.4 205 402 8.2 3 4 1.2 21 8 13.5 794.2 314.1 394.8 411.1 2288.5 658.6 153.4 16.1 60.6 6.4 76.3 8.0 79.4 8.4 442.1 46.3 Observed % 74 7.7 66 6.9 108 11.3 150 15.7 x2 41.1 .5 13.1 62.8 Grain source EC-RIM Extracellular ROS PM Cytoplasm Mitochondria Vesicles 661 9 8.1 61 53 2.1 Total Expected % ROS-RIM Vesicle-PM Total 6 3 2.6 6 928 966 9 4.9 2625 45.2 50.7 31.5 4943.5 127.3 13.2 9.8 1.0 6.1 0.6 955 236 24.7 179 18.8 101 10.6 41 4.3 955 100 96.1 21.0 848.7* 199.7* 1283.0 328 51.3 100 The source compartments (left column) were denned as those structures seen in the ellipsoid region (mitochondria, smooth membrane vesicles and cisternae, cytoplasm, and plasma membrane) as well as the region surrounding the ellipsoid (extracellular, ROS) which could also contribute grains to the ellipsoid region. The grain compartments (columns) consisted of the external rim regions (EC, ROS) as well as mitochondria and cytoplasm. The remaining grain compartments consisted of membranous components identified asjunctional items with cytoplasm. Thus, they were identified only in association with surrounding cytoplasm. In each case, the structure was identified by use of a circle with a 1 HD radius. An exception was the plasma membrane grain compartment which included grains within 1 HD inside or outside of the compartment. The center of the circle including an entire filamentous grain was taken as the locus of that grain. The source: grain compartment matrix ("crossfire matrix") was generated using the validated masks for 3H published as Figure 4 in Salpeter et al.38 These masks show the expected contribution of radioactivity in each source to silver grains in each grain compartment if radioactivity were uniformly distributed. Because mask sources rarely fell over vesicles or plasma membrane which were of principal concern in our analysis, sampling was improved by separate analysis of vesicles alone followed by normalization of the expected grains from these sources to the entire matrix by a procedure devised by Salpeter et al.38 The column totals represent the expected distribution of silver grains over the grain compartments. Sources for individual real silver grains were of course unknown. Observed grains were tabulated using the same definitions for grain compartments as was used in generating the matrix. Note that the high x 2 and low probability indicate significant differences between observed and expected. The asterisks indicate some of the individual grain compartments which contribute to the high x 2 . EC: extracellular; PM: plasma membrane; ROS: rod outer segment; RIM: rim region immediately outside of the ellipsoid. See Figure 4 for further definition of the cellular regions analyzed. density values necessary to account for the observed distribution of silver grains. Retinas of 10 adults were embedded in Lowicryl K4M31 (Polysciences) and were examined by immunocytochemistry. Eleven retinas from tadpole, juvenile, and adult Xenopus were also embedded in BSA30 and comparably studied.40 Thin sections were labeled sequentially for 15 min with the following: (1) biotinyl-rabbit antifrog opsin, 0.1 to 0.4 mg/ml (affinity purified F(ab')2 EM Immunocytochemistry Xenopus adults were killed under dim red light after 11-12 hr of dark adaptation or after light adaptation. Table 2. Computed grain distribution compared to observed using x 2 distribution Computed distribution Observed x2 EC-RIM ROSRIM PM Cytoplasm Mitochondria Mitochondriacytoplasm Vesiclecytoplasm VesiclePM 73.8 74 0.0007 65.9 66 0.0001 109.8 108 0.0312 153.9 150 0.0971 238.7 236 0.0299 168.7 179 0.6306 104.2 101 0.1031 40.0 41 0.0199 Using a computer program developed by Besharse and Schmidt similar to that described by Land and Salpeter (1978), we estimated the density of radioactivity in each source (source density) which would be necessary to yield a real distribution of silver grains like that observed. The basis of the computer program is a x 2 minimization routine which modifies the rows of the matrix with multipliers (source density values) until the hypothetical grains in the grain compartments yield a distribution which gives the lowest possible x 2 when compared to the real distribution of silver grains. Our computer program differs from that of Land and Salpeter (appendix to reference 38) principally Total 955 955 0.9126 in that it is written in BASIC and can be used on the Apple II microcomputers. In a validation test, both the Land and Salpeter program and our own program gave the same answers to a series of sample problems. Note at the bottom of Table 2 that in the final iteration of the program the X2 values are minimal and the new computed distribution is not significantly different from the observed distribution. The source density values can be regarded as multipliers for each source which when applied to the grain compartments alter the expected distribution. Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1390 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / Ocrober 1985 Table 3. Computed source densities for source compartments Source Computed source density* Extracellular ROS PM Cytoplasm Mitochondria Vesicles 0.031 ±0.02 0.114 ±0.04 1.243 ±0.68 0.412 ±0.04 0.044 ±0.01 6.903 ± 0.68 Arecrf % Activity^ Relative specific activity§ 18.8 6.6 1.0 19.6 53.1 0.9 3.0 3.9 6.7 41.7 12.1 32.6 0.162 0.589 6.433 2.132 0.228 35.73 % * Computer generated source densities which yield the computed grain distribution shown in Table 2. Source densities were generated as described in Table 2, and are reported as the density value ± a probable error estimate calculated according to the procedure of Salpeter et al.43 t Stereometric estimates derived from the grain source row totals in Table I. $ This is calculated as the source density X total hypothetical grains in source compartment divided by total observed grains. § This is calculated as % activity/% area. fragment or biotinyl-IgG 22 ; (2) biotinyl-F(ab')2 of sheep antirabbit F(ab') 2 , 0.1 mg/ml (affinity purified IgG); and (3) avidin-ferritin (0.03 mg/ml in 0.1 M TrisHC1 pH 7.4.30-41 The rabbit antifrog opsin sera used in this study were the same as the sera designated serum 1 and serum 2 previously.29 Data presented in Table 4 and Figures 5-8 are derived from duplicate experiments using antiserum 2. Figure 9 was from an experiment using antiserum 1. Sections were stained with aqueous uranyl acetate and bismuth subnitrate42 after the immunocytochemical sequence. Controls consisted of replacement of the first-stage antibody with biotinylpreimmune or biotinyl-nonimmune rabbit F(ab') 2 , or biotinyl IgG for the corresponding antiopsin antibody. Vol. 26 Preimmune sera were obtained from rabbits subsequently immunized; nonimmune sera were obtained from unimmunized rabbits. Quantitative distribution of labeling was estimated morphometrically by point counting of duplicate experiments of retinas obtained after 1 hr of light exposure of images magnified at least XI 00,000. The specific data for Table 4 were obtained by random sampling to eliminate observer bias in selection of labeled areas. On duplicate grids from duplicated experiments, a centrally located section was sampled by use of a random number table (from 1 to 5) to select the cell by counting from the left grid bar. If the cell thus selected was a cone, the adjacent rod was photographed. Images were recorded on 35-mm film and each roll contained an image of a calibration grid to determine the magnification. The values for outer segment plasma membrane labeling were obtained from all rods in the sections which were oriented so that the plane of section had passed through the connecting cilium and its projecting microtubules in the ROS. The ciliary shaft separated the plasma membrane from the disks (Fig. 7a). Linear labeling densities along the rod outer and inner segment plasma membrane were obtained by projecting the image on to a lattice with a Bellco plaque viewer (Bellco; Vineland NJ) and sampling at intersections of the membrane and the lattice lines. A second lattice was superimposed and aligned along the plasma membrane so that the intersections with the first lattice were centered in the square of the second lattice. The number of ferritin particles within a square were counted and counts from five squares were collected to determine the mean density by the equation: 7VF(c) = NP/dPc Table 4. Immunocytochemical labeling densities of the inner and outer segment membranes of rod photoreceptors Ferritins/fim2 Ferritins/nm Cellular site Antibody Biotinyl-affinity purified antiopsin No. 2 Biotinylnonimmune IgG ROS plasma membrane Ellipsoid plasma membrane* Myoid plasma membrane 55 ±5 N = 17 6 ±2 N = 9 7 ± 1 N = 20 0.5 ± 0.2 N = 15 0.9 ± 0.2 N = 17 0.5 ± 0.2 N = 15 * In one experiment, sections of a retina from a dark-adapted tadpole were heavily labeled on the ellipsoid plasma membrane. All sections from those blocks labeled in that pattern but subsequent experiments with other animals could not reproduce this phenomenon. t The labeling density of Golgi membranes varied greatly in micrographs not selected at random (range 200 to 1000 ferritins//im2). All data in this table are means of duplicate experiments whose labeling densities were determined Rod outer segments 1189±81 N = 9 5 ± 0.5 N = 16 Mitochondria Golgi membranes SER Interphotoreceptor matrix 8± 2 N = 19 353 ± 29f N = 13 38 ± 10* N = 10 8±4 N = 20 5± 1 N = 17 10 ± 2 N = 10 10 ± 2 N = 14 2 ±0.5 N = 16 by random sampling (see text for details). N = the number of micrographs counted to determine the mean and standard error. % See Mercurio and Holtzman43 for discussion of the relative contributions of membrane density surface area to the observed area density in the subellipsoid SER. When corrected for the high membrane density of this closely apposed set of membranes, the labeling density approaches background. Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. No. 10 where NF is the sum of ferritin counts in each of the small squares, d is the lattice spacing of the small square (in micrometers corresponding to the magnification of the image) and Pc is the number of lattice points intersecting the plasma membrane. These densities were summed from all micrographs (N = 9 to 20) to determine the mean density ± standard error (SE) of the sample analyzed. Area densities of the Golgi, smooth endoplasmic reticulum (SER), mitochondria, and ROS were obtained as previously described.30 Results EM Autoradiography At 2 hr 15 min postinjection, we found that considerable radioactivity was still contained in the myoid region of the inner segment in close association with the Golgi complex and, to a lesser extent, with the rough endoplasmic reticulum (RER) (Fig. 2). Radiolabeled proteins were also observed in the ellipsoid and clustered beneath the periciliary ridge complex at the base of the connecting cilium (Figs. 3, 4). By 3 or 4 hr after injection, extensive transfer of radioactivity from inner to outer segment had occurred. This time course and path of transport of radiolabeled rod proteins in X. laevis retinas is similar to earlier observations of renewal of ROS proteins in R. pipiens.ll-l9A3M The principal question that we wished to answer by detailed analysis of sources of radioactive decay was whether or not radioactivity in the ellipsoid region of the inner segment, destined for ultimate incorporation into ROS disks, was associated with the rich array of vesicular cytomembranes observed in that region. Qualitative evaluation, particularly of the vesicle-rich periciliary region (Fig. 3), suggested association of radioactivity with abundant vesicles seen there. A small amount of radioactivity was already associated with ROS basal disks by 2 hr and 15 min of incorporation. The 2 hr or 2 hr and 15 min time point was chosen for detailed analysis because radioactivity was abundant in the ellipsoid and transfer to the ROS had just begun. Within the ellipsoid, autoradiographic grains were associated with vesicles between mitochondria and vesicles clustered about the base of the connecting cilium (Fig. 3). However, most grains were associated with other organelles (see Fig. 4 and Table 1, row labeled Observed) including cytoplasm, mitochondria, and the perimitochondrial space. Because of the close spatial packing of vesicular membranes and mitochondria within the ellipsoid (Fig. 4) and the low level of resolution attainable with autoradiography relative to the size of vesicles, we used quantitative procedures to assess the sources of radioactive decay. In a preliminary 1391 analysis, we used the probability circle method 39 in which an expected distribution of silver grains based on the hypothesis of uniform labeling was generated. We found that the actual distribution of silver grains differed from a uniform distribution with a high degree of probability (P < 0.001). This analysis showed that the compartments with major deviations from a uniform distribution were mitochondria, vesicles, and plasma membrane. The latter compartments, of necessity, also included the immediately adjacent cytoplasm. The mitochondria contained far less label than would have been expected if the distribution were uniform, whereas both vesicles and plasma membrane contained far more label. The study of both paraformaldehyde- and glutaraldehyde-fixed material in this and the subsequent analysis (see below) led to the same conclusion regarding the localization of radioactive sources. Although we are unable to rule out a low level of nonspecific association of radioactivity with tissue due to glutaraldehyde fixation,27 our data shows this had little if any effect on the localization. In order to estimate the relative amount of radioactivity contained in identifiable organelles rather than in grain compartments, we utilized the hypothetical grain distribution method.36"38 We obtained estimates of relative source densities within organelles which would lead to a given pattern of real grains over grain compartments. 38 The analysis took three stages. First, a hypothetical source:grain compartment matrix was generated using overlays designed by Salpeter38 based on knowledge of the resolution of the autoradiographic technique. This matrix provided estimates of the silver grains to be expected over defined grain compartments if each organelle contributed to the distribution in proportion to its fractional area in the collection of micrographs (see definitions of source and grain compartments in Fig. 4 and Table 1). Second, real silver grains were tabulated over grain compartments using the same definitions that were used in generating the hypothetical matrix. Comparison of the expected totals in each grain compartment with the observed totals, although formally analogous to the method used by Williams,39 differs in its definition of compartments. The hypothetical matrix and expected distribution under the assumption of uniform distribution of sources are compared to the observed distribution of silver grains in Table 1. Expected and observed distributions differed from each other significantly (P < 0.001). As in our preliminary analyses using the probability circle method, the major deviations from uniformity were attributable to the low level of mitochondrial and high level of membrane compartment labeling compared to an expected distribution if sources of radiolabel were Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 OS M e m »- N Fig. 2. Low power autoradiograms illustrating the overall distribution of radioactivity in inner segments 2 hr and 15 min after injection of radioactive amino acids (2 hr in light). This figure and subsequent figures were obtained from retinas fixed in a glutaraldehyde-paraformaldehyde mixture, a, Autoradiogram developed with Agfa-Gevaert fine grain developer of entire rod inner segment from a postmetamorphic juvenile retina. The silver grains are localized predominantly over the Golgi apparatus and to a lesser extent over RER at this period of incorporation. C: connecting cilium; G: Golgi apparatus; M: mitochondria of the ellipsoid; N: nucleus; ROS: rod outer segment (bar = 1 ^m; X6,500). b, Autoradiogram developed with Phenidon-ascorbic acid of myoid region of a similar inner segment from a tadpole retina. The radiolabeled protein is concentrated in the Golgi apparatus which extends from the perinuclear region to the base of the ellipsoid. G: Golgi apparatus; M: mitochondria; N: nucleus (bar = 1 (im; X 16,000). Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 Fig. 3. High power autoradiograms illustrating the distribution of radioactivity in the rod periciliary region, 2 hr and 15 min after injection of radioactive amino acids into a postmetamorphic X. laevis juvenile. Developed with Agfa-Gevaert fine grain developer, a, Section through the connecting cilium showing association of silver grains with vesicular membrane profiles and basal rod outer segment (ROS) disks and a lack of grains over mitochondria (M). Some vesicles are juxtaposed to the grooves (G) of the periciliary ridge complex. The connecting cilium (C) joins the ROS to the inner segment. Vesicles (V) are clustered beneath the basal body and are scattered and compressed among the densely packed mitochondria. Radiolabeled protein is associated with the vesicles and some has passed to the ROS along the basal disks at this time—2 hr (see Table 1) (bar = I M m; X27,000). b, Section through the periciliary region as evidenced by presence of an associated centriole of the basal body complex (B) and the ridge (R) and scalloped grooves (G) of the periciliary ridge complex. At this time of incorporation (2 hr), silver grains are associated with vesicles (V) and periciliary grooves. A vesicle is captured during apparent fusion with the base of a groove (arrow) and may represent either delivery of opsin by exocytosis or endocytosis of plasma membrane (bar - 1 (jm; X30,500). 1393 VESICULAR TRANSPORT OF OPSIN / Papermasrer er al. ROS G .R M uniform. The major contributor to the high x 2 value was the vesicle-cytoplasm compartment which contained more than 10 times the number expected on the basis of a uniform distribution (Table 1). Finally, in our third stage of analysis, a computer program designed to minimize x 2 by varying the source density values was used to fit the hypothetical distribution in grain compartments so that the calculated compartment totals corresponded to the observed grain totals (Table 2). The optimized source density values ROS R 0* V B are listed in Table 3 in units related to the numbers of hypothetical grains in the matrix and as a percent of total radioactivity. For comparison, the area fraction of each source and relative specific activity (% activity/ % area) are also presented. Table 2 also includes the minimized x 2 values which correspond to the optimized source densities. The results in Tables 2 and 3 were derived by simulation but provide our best estimate of the radioactivity contained in each source. The data suggest that Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1394 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Ocrober 1985 Vol. 26 •**TT Fig. 4. Autoradiogram of the ellipsoid region of a tadpole retina after 2 hr and 15 min of incorporation of radiolabeled amino acids. Developed with Kodak D-19. This figure illustrates some of the grain compartments used for quantitative analysis. The dark line outlines the region defined as ellipsoid. The dotted line defines a RIM region 5 HD units wide around the ellipsoid. This region could contain silver grains which originate from sources of decay in the ellipsoid. It constitutes both the potential source compartment and a grain compartment. This RIM region is also heterogeneous in that it contains myoid, ROS, and extracellular components. In Table 1, it constitutes the source compartments named myoid, extracellular, and ROS and the grain compartments named extracellular-RIM and ROS-RIM. Other source and grain compartments are denned in Table I. Open arrows indicate some of the intermitochondrial vesicles in the ellipsoid. G: groove of the periciliary ridge complex; R: ridge (bar = I \im\ X20,000). Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. vesicle membranes contained about 30% of the radioactivity, and had a specific activity 17 times higher than surrounding cytoplasm. In contrast, we estimated that mitochondria, which occupy 64% of the area,f contained only 12% of the radioactivity and had a specific activity of 0.23. The source compartment labeled "plasma membrane" also exhibited a high specific activity. Although not considered as a separate source, much of this radioactivity appeared to be associated with the apical plasma membrane of the inner segment. However, close apposition of apical plasma membrane and lightly labeled basal ROS disks in most micrographs made it virtually impossible to distinguish the relative contributions of the two adjacent membranes as sources. EM Immunocytochemistry The other eye from each tadpole was embedded in glutaraldehyde cross-linked bovine serum albumin to localize immunoreactive opsin in the inner segment. Because of the low contrast in BSA-embedded retinas, we continued this investigation using Lowicryl K4M (Polysciences) because of its superior tissue contrast.31'32 The enhanced contrast readily demonstrated opsin bearing sites in Golgi and on vesicles in the ellipsoid region. Myoid region: The Golgi of X. laevis rod photoreceptors is axially oriented and may be semicylindrical or wedge-shaped since longitudinal sections occasionally demonstrated paired sets of Golgi membranes on each side of the myoid while oblique sections generated crescentic or V-shaped profiles (Figs. 5a-c). Antiopsin binding to the Golgi zone was easily appreciated even at low magnification as a result of the enhancement of labeling density by the three-stage technique (Fig. 5a). Labeling of Golgi membranes was nearly confluent, an indication that opsin may be highly concentrated at this site prior to its transport toward the ROS. Variation of labeling within the Golgi apparatus was considerable, however (Fig. 5b). Although a mean labeling density could be obtained by point counting (353 ± 29), the labeling densities spanned a much larger range than the ROS. Some domains of the Golgi apparatus were as low as 200 ferritins/Mm2 while other areas approached ROS in labeling density (Fig. 5a; Table 4). Surrounding the Golgi membranes was an ill-defined zone from which ribosomes were excluded (Figs. 5ac). This zone was invariably unlabeled by antiopsin so that the Golgi membrane profiles were well demarcated f This stereometric estimate is obtained from Table 1 by omitting area contributions from ROS and the extracellular compartments and computing the percent area of mitochondria in the ellipsoid proper. 1395 from the adjacent RER. Control sections labeled with biotinyl nonimmune IgG, biotinyl antilgG, and avidinferritin were negligibly labeled (Fig. 5 c). We have previously observed dense labeling of Golgi zones of frog photoreceptors with these antibodies applied to BSAembedded tissues.29 Between the Golgi and the mitochondria, smooth endoplasmic reticulum membranes are closely packed. Antiopsin binding was negligible over these membranes and approached background levels (Fig. 5d; Table 4). These unlabeled membranes were also shown to be inactive in incorporation of radiolabeled amino acids into protein and of radiolabeled choline and glycerol into glycerolipids by Mercurio and Holtzman.43 Profiles of closely packed membranes of unlabeled smooth endoplasmic reticulum were also found, occasionally, in a cytoplasmic channel that extended between the mitochondria of the ellipsoid from the myoid toward the periciliary ridge complex (Fig. 4). Ellipsoid region: Vesicular profiles labeled by antiopsin were observed not only within the channel that spanned the ellipsoid but also between the closely packed mitochondria (Fig. 6). Vesicles were not more commonly seen in the channel than outside it between mitochondria. These vesicular profiles usually corresponded in size to those seen in epon sections and may also represent cross-sections of serpentine cisternae. There appeared to be some artifactual expansion of some of the vesicles during Lowicryl embedding, however. Many of the vesicles were highly labeled but some were unlabeled or were labeled by antiopsin heterogeneously. Some were slightly labeled, others were eccentrically labeled, and a few were confluently labeled about the circumference of the vesicle in a pattern consistent with the binding of antibody on the cut edge of the vesicle membrane. Some labeling appeared within the vesicle interior, probably as a consequence of a tangential section of the vesicle's cytoplasmic surface or of labeling of the interior of the vesicle membrane. Inspection of the figures revealed that vesicle membrane labeling density varies from sparse (ca 7/^m) to confluent (ca 55//um) on labeled vesicles. Because of this heterogeneity and for reasons based on the unique geometry of vesicles (see Discussion), we were unable to compare the vesicle membrane labeling density directly to the labeling density of the adjacent RIS and ROS plasma membranes that are detailed in Table 4. Stereo images of the Golgi and ellipsoid regions do not indicate significant penetration of the surface of the Lowicryl section. Thus labeling densities are not confounded by superposition of ferritins from within the depth of the section. Poor tissue contrast in BSA-embedded retinas and compression caused by dehydration obscured potential distinction of collapsed vesicles between mitochon- Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1396 INVESTIGATIVE OPHTHALMOLOGY G VISUAL.SCIENCE / October 1985 Vol. 26 RER N RER MC Fig, 5. Immunocytochemical localization of opsin in the myoid region of the inner segment of A', laevis rod photoreceptors embedded in Lowicryl K4M. These figures and Figures 6-8 are of Xenopus laevis adults killed I hr after light onset after entrainment to a 12 hr light: 12 hr dark cycle of light exposure. The three-stage label consisted of affinity purified biotinyl-rabbit IgG of antifrog opsin, biotinyl-sheep F(ab')2 of antirabbit F(ab') 2 , and avidin-ferritin. The three-stage label enhances the density of bound ferritins without increasing background. An average of 5-7 ferritins are clustered on each bound antiopsin. a, The closely stacked Golgi membranes (G) near the nucleus (N) are obscured by the dense label of antiopsin-ferritin complexes bound on the section surface. The perinuclear plasma membrane (arrows) and adjacent Miiller cell (MC) with its processes (P) are unlabeled (bar = 1 /im; original magnification, X20,000). b, This axially arrayed Golgi is labeled at a density (449/Vm2) that is near the mean density of several rods (see Table 4). The labeling is not uniform over the entire Golgi, however. The Golgi zone is separated from the surrounding RER by a low-contrast border which is relatively unlabeled (bar = 0.5 ftm; original magnification, X 39,000). dria.40 Nonetheless, antiopsin labeling in the ellipsoid of BSA embedded rods was largely confined to the intermitochondrial space (see Fig. 9 in ref. 40). The improvement in tissue contrast of Lowicryl-embedded tissues more readily permitted interpretation of the labeling in this area. Mitochondria] labeling density approached background levels (Table 4). Control sections demonstrated no significant nonspecific binding of second or third stage reagents (Fig. 7c; Table 4). Only scant labeling was noted along the lateral plasma membrane of the inner segment (Figs. 5a, 6; Table 4). Sections including plasma membrane alongside the actin filament bundles that extend from the calycal processes toward the myoid45 were not labeled to a greater extent. Between the actin bundles, the plasma membrane was labeled by rare ferritin clusters (Figs. 5-7), a result which parallels studies of R. pipiens retinas labeled by immersion.46 Periciliary ridge complex: The plasma membrane surrounding the base of the connecting cilium of R. pipiens rods and cones is highly folded into an array of nine ridges and grooves to form a domain termed the periciliary ridge complex (PRC).47 X. laevis rods appear to have a comparable domain. Both longitudinal sections and cross-sections of the ellipsoid revealed densely labeled vesicles beneath the PRC Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 VESICULAR, TRANSPORT OF OPSIN / Papermosrer er ol. 1397 M SER RER Fig. 5c. Control section labeled with biotinyl-nonimmune IgG, biotinyl-sheep F(ab')2 of antirabbit F(ab')2, and avidin-ferritin. The Golgi membranes (G) are unlabeled. A rare antibody-ferritin complex is circled (bar = 0.5 fim; original magnification, X32,000). d, Near the junction of the myoid and ellipsoid, smooth endoplasmic reticulum (SER) membranes form close-packed arrays. SER membranes and the adjacent mitochondria (M) are virtually unlabeled. The very tow level of labeling of the lateral plasma membrane (arrow) indicates that this membrane apparently does not participate as a major pathway for transport of opsin to the outer segment (see Table 4) (bar = 1 fim; original magnification, X28,000). Inset, Epon-embedded rod cell illustrating the conventional appearance of the subellipsoid SER (bracket) (bar = I jim; original magnification, X 14,000). grooves (Figs. 7-9). Vesicular profiles were observed in all planes of section which indicated that their shape was predominantly spherical in this region and probably did not arise from an elaborately folded cisternal array (compare, for example, the appearance of the subellipsoid SER cisternae, Fig. 5d, inset). In an occasional section, a vesicle was captured in apparent fusion with the base of the groove of the PRC (Fig. 8). Microtubules radiated from the basal body region into the ellipsoid but vesicles were not obviously aligned along them. Despite their close proximity to the plasma membrane of the grooves and ridges of the PRC, the vesicles were excluded from the volume immediately surrounding the accessory centriole and basal body of the connecting cilium (Fig. 7a-c). The area of vesiclefree cytoplasm measured 0.1-0.2 jum in diameter. The high labeling density of these periciliary vesicles was difficult to quantify because the diameter of the antibody-ferritin clusters precluded assignment of the bound ferritins to one vesicle or its neighbor (see Discussion). The labeling density clearly exceeded the labeling of the adjacent RIS plasma membrane which could be easily quantitated (Fig. 7; Table 4). On some vesicles, the labeling density approached the density of the rod inner segment plasma membrane. The basal plasma membrane of the connecting cilium usually was unlabeled. In cross-sections, the cilium was occasionally labeled on its plasma membrane (Fig. 9). This may be an indication of transport along the plasma membrane of the cilium to the outer segment in axial lanes47'48 and is the subject of further study. Beyond the basal cilium, the plasma membrane of the distal cilium and rod outer segment was confluently labeled at levels 10-fold above the adjacent inner segment plasma membrane (Fig. 7a; Table 4). The inner segment plasma membrane was labeled at levels which approached background (Figs. 5-7; Table 4). This emphasizes the extraordinary polarity of opsin distribution in the rod photoreceptor plasma membrane despite its continuity across the connecting cilium. Discussion Our study provides direct evidence that specific vesicular membranes vectorially transport newly synthesized opsin across the large intracellular space from the Golgi zone to the periciliary ridge complex. Quantitative analysis of sources of radiolabeled protein decay indicated that the most heavily labeled structures in the ellipsoid were the membranous vesicles. Despite Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1398 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / Ocrober 1985 Vol. 26 M Fig. 6. Longitudinal section of the ellipsoid region (Lowicryl K.4M embeddment) labeled with antiopsin-ferritin complexes as in Figure 5. An oblique channel passing between the mitochondria (M) to the base of the connecting cilium is often observed in sections through the center of the cell. Free ribosomes, rough endoplasmic reticulum cisternae, microtubules, stacked smooth endoplasmic reticulum, short tubular cisternae, and vesicles are often seen in this channel. This micrograph illustrates an unusual degree of clustering of heavily labeled vesicles in the channel (V). Most vesicles are packed between the closely apposed mitochondria (arrows). Occasionally, labeled vesicles are found beneath the lateral plasma membrane (open arrow). The labeling density of the lateral plasma membrane varies (arrowheads) but does not approach the level of vesicle or Golgi (G) labeling (bar = 1 ^m; original magnification, X31,000). Inset, Higher magnification image of ellipsoidal vesicles illustrating the labeling of the vesicle margins (original magnification, X74,000). Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 1399 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. M M M Fig. 7. Longitudinal section of the connecting cilium joining the inner and outer segments of X. laevis rod photoreceptors. a, Section of Lowicryl-embedded retina labeled with antiopsin-ferritin complexes as in Figure 5. The tightly clustered vesicles (V) beneath the cilium (C) as well as the rod outer segment (ROS) disks and ROS plasma membrane (arrowhead) are confluently labeled. Mitochondria (M) and the lateral plasma membrane of the inner segment (arrow) are unlabeled (bar = I ^m; X48,000). b, Section of epon-embedded retina. Vesicles clustered beneath the cilium are comparable in size and distribution to those seen after Lowicryl embeddment (bar = I nm; x35,000). c, Control section labeled as in Figure 5a inset. Ferritin density is insignificant (bar = 1 fim; X30,000). the small tissue volume they occupied (Tables 1-3), they exhibited a relative specific activity far greater than any other organelle. Two points should be considered regarding the reliability of our source density estimates. First, similar results were obtained in three additional analyses (data not shown). Second, error estimates for the source density values using the procedure of SaU peter et al38 indicated that the individual density values in the analyses were reliable (see Table 3, column 1). We found that a substantial proportion (41.7%) of the total radioactivity was associated with cytoplasm. In the myoid region, a higher proportion of the total radioactivity was associated with rough endoplasmic reticulum and Golgi apparatus (Fig. 1). The silver Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1400 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / October 1985 Vol. 26 pected. Previous studies21 have shown that most of incorporated label in the intermediate and microsomal membrane fractions (fractions 2 + 3) from whole retina was contained in opsin but that cytosol fractions (fraction 4) contained nonopsin radiolabeled proteins. Overall, the high relative specific activity of the vesicle compartment (Table 3, column 4) highlights the preponderant role of this set of membranes in transport of newly synthesized opsin. V B M Fig. 8. Oblique section passing through the periciliary ridge complex (PRC). A vesicle (arrow) is captured while in apparent fusion with the groove of the PRC. Some labeled vesicles between mitochondria (arrowhead) may be as confluently labeled as the juxtaciliary vesicles, others are eccentrically labeled. Vesicles are excluded from the cytoplasm surrounding the accessory centriole of the basal body. B: accessory centriole of the basal body; G: PRC groove; R: PRC ridge; M; mitochondria; ROS: rod outer segment (bar = 0.5 fim; original magnification, X45,000). grains in the "myoid-RIM" region (see Fig. 4 for definitions) were included largely in the "cytoplasm" grain compartment. Because many grains probably originated from disintegrations in adjacent heavily labeled RER and Golgi compartments, their inclusion in "cytoplasm" would cause us to overestimate the labeling of this compartment in the ellipsoid. The assignment of radioactive sources to cytoplasm may also reflect an underestimate of the vesicle-related radioactivity that would occur because the large filamentous silver grains obscure underlying ultrastructure. The small size of the vesicles would lead to an erroneous association of grains with cytoplasm and an overestimation of the cytoplasmic grain density. Despite these uncertainties, a significant level of cytoplasmic labeling would be ex- To determine if the radiolabeled vesicle population also contained opsin, we conducted parallel immunocytochemical localization studies with antiopsin antibodies. Binding of these antibodies was restricted to vesicular and cisternal profiles in intermitochondrial spaces and the vesicle-rich region adjacent to the base of the connecting cilium (Figs. 6-9; Table 4). In a more recent study, combined freeze-fracture and immunocytochemical labeling demonstrated binding of sheep antibovine opsin to the interiors of ellipsoidal vesicles.48 The Golgi apparatus was labeled prominantly by antiopsin. The.labeling density was less than the density over ROS but much greater than the adjacent RER (Figs. 5a-c). To the extent that immunocytochemical labeling density reflects antigen density—and not just antigen exposure—the greater labeling of Golgi apparatus embedded in both Lowicryl and albumin suggests that newly synthesized opsin is concentrated in Golgi membranes prior to its vectorial transport to the periciliary ridge complex. This result with Xenopus retinas parallels our earlier localization of opsin in the Golgi of Rana rod photoreceptors.29 The Golgi apparatus also became heavily labeled with radioactive proteins within 1 hr of incorporation. The time-course of passage of newly synthesized protein from RER to Golgi evaluated by EM autoradiography was comparable in these X. laevis retinas and the Rana retinas studied by Young and Droz18 and Hall et al.17 The structure of bovine opsin's oligosaccharides and autoradiographic studies of radiolabeled sugar incorporation in frog retinas suggested that the Golgi probably completes the processing of the oligosaccharides by addition of N-acetyl-glucosamine to the nonreducing terminal.49"51 Amphibian opsins are also glycosylated and comparable synthetic steps may occur in their photoreceptors. The high density of opsin in the Golgi zone revealed by immunocytochemistry also suggests that it may serve as a center for concentration of opsin prior to its distribution to the outer segment. Comparable functions for the Golgi apparatus have been proposed in studies of virus biosynthesis.7'8 The absence of opsin in soluble cytosol fractions21 and the similar size of vesicular intramembranous particles (IMP) and of ROS IMP in freeze fracture studies23 suggested that the vesicles contained opsin. The IMP density of the vesicles was only half that of the ROS Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. * * • < 1401 . V Fig. 9. Cross-section of the basal portion of a rod photoreceptor connecting cilium (C) as it arises from the periciliary ridge complex (embedded in Lowicryl K4M and labeled as in Figure 5 with antiopsin serum 1). The animal was killed 3 hx after light onset. The ridges (R) and grooves (G) form a deep invagination about the base of the cilium. Vesicles (V) are tightly clustered in this region and are highly labeled by antibodyferritin complexes on their margins. Occasional vesicles are labeled on the vesicle's interior (open arrow). Antibodies are also bound on the plasma membranes of the ridges and grooves. A few complexes label the plasma membrane of the cross-sectioned basal cilium alongside the microtubules. This may be an indication of a final pathway for transport of opsin from the inner to the outer segment along the ciliary plasma membrane (bar = 1 fim; original magnification, X51,000). disks which may be an indication that a two-fold concentration of opsin and destruction or retrieval of an equivalent amount of IMP-free membrane occurs. Preliminary evidence for retrieval of membrane has been demonstrated in short-term retina cultures that were exposed to horseradish peroxidase. Some of the vesicles in the ellipsoid region became labeled with endocytosed peroxidase. 5253 Some of the unlabeled vesicles of the ellipsoid region revealed in both the autoradiographic and immunocytochemical portions of this study may represent this population of endocytic vesicles (Figs. 4, 6-9). It should be noted that we did not identify separate subpopulations of vesicles. In the autoradiographic analysis, the inclusion of an unlabeled subset of vesicles would cause us to underestimate the true specific activity of the vesicles transporting newly synthesized opsin. In the immunocytochemical studies, many of the intermitochondrial vesicles in the ellipsoid region and the vesicles clustered beneath the periciliary ridge complex appeared as poorly demarcated pale oval structures covered partially by an area of clustered label (Fig. 9, open arrow). Other vesicles were labeled on their circular margins (Figs. 7-9). This variation in labeling distribution and vesicle appearance would be expected because the vesicle's small diameter approached the thickness of the thin sections. If some of the vesicles were tangentially sectioned so that their membranes were barely exposed on the thin section's surface, the membrane-bound label would appear to fall over the vesicle's interior. For this reason, we cannot expect and did not observe label exclusively associated with the margins of circular or elongated membrane vesicles and cisternae cut in cross section. Use of thin-sectioned retinas embedded in hydrophilic media eliminates the impermeable plasma membranes as a barrier to labeling by antibody-ferritin complexes by exposing intracellular antigenic sites di- Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1402 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / Ocrober 1985 rectly on the section surface. Quantitative comparison of labeling densities of well-ordered intracellular membranes exposed on the surface of such sections may be informative. Such comparisons of labeling densities of rod and cone photoreceptor outer segments readily demonstrated different degrees of immunocytochemical cross-reactivity among these cells.29 We have applied this quantitative approach confidently to certain domains of the rod inner segment (Table 4). In the less homogeneous areas of the inner segment, however, the geometric factors just discussed, as well as the actual density of an antigen, may contribute significantly to the observed labeling density. Interpretation of labeling densities on these domains is therefore of much more limited value so that we could not determine an apparent labeling density of the periciliary vesicles with the same degree of confidence. The vesicles in the ellipsoid and beneath the periciliary ridge complex were labeled to a variable extent. Some were unlabeled, others were eccentrically or lightly labeled, while many were confluently labeled (Figures 6 and 9). If only labeled vesicles are considered, the densities ranged approximately from 7 to 60 ferritins//*m of vesicle membrane length. The most confluently labeled vesicles clearly exceeded the labeling densities of the adjacent RIS plasma membranes and approached the appearance of the labeling of the ROS plasma membrane. Since some of the vesicles are closely packed, however, the apposed membranes may contribute to this appearance. This technique is clearly unable to establish directly the absolute antigen density of the vesicles. Identification of opsin as a major vesicle component, however, increases the likelihood that the IMP observed in the vesicles by freeze fracture are opsin clusters. Freeze fracture readily permitted an evaluation of particle density in different domains of the cell and indicated that the opsin density of vesicles was half that of the ROS disks.23 Comparisons of labeling density over other domains of the inner segment are offered as an indication of the variation of labeling in repeated experiments in order to document the reproducibility of the observations and to highlight the high labeling density in the Golgi apparatus and the clear demarcation of the boundaries of opsin insertion in the plasma membrane (Table 4). The high labeling density of the Golgi apparatus and the periciliary vesicles and the large numbers of these vesicles clustered beneath the PRC suggests that opsin transport from sites of synthesis in the rough endoplasmic reticulum to the outer segment is associated with high steady-state concentration of the protein at these two sites. To what extent the absolute opsin density in these sites is subject to physiologic stimuli is not yet known. We attempted to evaluate the role of light Vol. 26 to determine if the periciliary vesicle population might become depleted by light exposure. Previously, it had been observed that most of the new disks formed during a day were assembled in the first 8 hr.15 We did not observe any large-scale depletion of the vesicle population at any time in the circadian cycle (cf Figs. 7-9), even after prolonged exposure to darkness and short exposure to light, a condition which favors a major increase in disk assembly.54 Stimulated by these observations, we have begun intensive study of the periciliary region of the cell tc identify the cellular components involved in the last stages of opsin delivery to the outer segment. We have observed that the plasma membrane of proximal portions of both amphibian and rat connecting cilia were nearly unlabeled by antiopsin when compared to the dense label of membranes of the distal portions of the cilium or the adjacent vesicles beneath the cilium.40 High resolution scanning electron micrographs of the apical plasma membrane of the frog rod and cone inner segment revealed an extraordinary array of nine ridges and grooves surrounding the connecting cilium which was named the periciliary ridge complex.47 It may be a site for insertion of opsin-bearing vesicles as they terminate their passage through the inner segment. Both the autoradiographic and immunocytochemical results indicate that the vesicles may fuse with the base of the groove (Figs. 3b, 8). Low density labeling of the base of the cilium contrasts sharply with the nearly confluent labeling of the distal ciliary plasma membrane (Fig. 7a). Consequently, we have proposed that the base of the cilium may be part of a one-way gate forming the boundary between the outer and inner segment—permitting passage of opsin to the ROS but restricting back diffusion on to the lateral plasma membrane of the inner segment.23'40'46 The photoreceptor cell apparently has separated, both spatially and kinetically, the processes of membrane protein biosynthesis, transport and insertion. Our evidence that opsin laden vesicles transport opsin in rod photoreceptors raises several fundamental issues concerning the cell's mechanisms for intracellular vectorial transport of membrane proteins. Since the vesicle membranes are often confluently labeled and have high 10 nm IMP density, it is likely that the opsin content of these specific vesicles is quite high, perhaps as high as 50% of the density of opsin in the ROS disk. Does opsin contain, in some domain of the molecule, the specific information perceived by the inner segment as an "address" to direct it uniquely to one pole of the cell or does the vesicle contain other molecules with this function? What provides the "address" for the membrane proteins destined for the lateral and synaptic membrane proteins? If the Golgi apparatus is a center for sorting membrane molecules to their specific des- Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 No. 10 VESICULAR TRANSPORT OF OPSIN / Popermosrer er ol. tination, is the same Golgi apparatus used for biosynthesis and sorting of other photoreceptor membrane proteins as appears to be the case for cells infected simultaneously with two viruses?7 While these same questions can be asked of any polarized cell, the photoreceptor continues to provide an especially valuable object to probe them without perturbing its normal functions and renewal of its membranes. 12. 13. Key words: antibody, autoradiography, Golgi, immunocytochemistry, membrane biosynthesis, rhodopsin, vesicle, Xenopus 14. Acknowledgments 15. The authors are grateful to Dr. Beth Burnside for pointing out the possible role of inner segment vesicles from her early unpublished studies of monkey retinas. The authors thank Drs. Ewald Weibel and Jean Pierre Kraehenbuhl for helpful discussions of morphometry and immunocytochemistry, Dr. Frederick Schmidt for his efforts in writing the computer program for autoradiographic analysis, and Dr. Miriam Salpeter for advice on autoradiography and for running sample autoradiographic data sets from our study on her computer program to confirm the accuracy of our derived program. 16. 17. 18. 19. References 20. 1. Abrahamson DR and Rodewald R: Evidence for the sorting of endocytic vesicle contents during the receptor-mediated transport of IgG across the newborn rat intestine. J Cell Biol 91:270, 1981. 2. Kuhn LC and Kraehenbuhl JP: The membrane receptor for polymeric immunoglobulin is structurally related to secretory component. J Biol Chem 256:12490, 1981. 3. Steinman RM, Mellman IS, Muller WA, and Cohn ZA: Endocytosis and the recycling of plasma membrane. J Cell Biol 96:1, 1983. 4. Willingham MC and Pastan I: The receptosome: an intermediate organelle of receptor-mediated endocytosis in cultured fibroblasts. Cell 21:67, 1980. 5. Matlin K, Bainton DF, Pesonen M, Louvard D, Genty N, and Simons K: Transepithelial transport of a viral membrane glycoprotein implanted into the apical plasma membrane of MadinDarby canine kidney cells. I. Morphological evidence. J Cell Biol 97:627, 1983. 6. Pesonen M and Simons K: Transepithelial transport of a viral membrane glycoprotein implanted into the apical plasma membrane of Madin-Darby canine kidney cells. II. Immunological Quantitation. J Cell Biol 97:638, 1983. 7. Rindler MJ, Ivanov IE, Pleskin H, Rodriguez-Boulan E, and Sabatini D: Viral glycoproteins destined for apical or basolateral plasma membrane domains traverse the same Golgi apparatus during their intracellular transport in doubly infected MadinDarby canine kidney cells. J Cell Biol 98:1304, 1984. 8. Rothman JE, Fries E, Dunphy WG, and Urbani LJ: The Golgi apparatus, coated vesicles, and the sorting problem. Cold Spring Harbor Symp Quant Biol 46(Pt. 2):797, 1982. 9. Rodriguez-Boulan E and Sabatini.DD: Asymmetric budding of viruses in epithelial monolayers: a model system for study of epithelial polarity. Proc Natl Acad Sci USA 75:5071, 1978. 10. Rodriguez-Boulan EJ and Pendergast M: Polarized distribution of viral envelope glycoproteins in the plasma membrane of infected epithelial cells. Cell 20:45, 1980. 11. Rodriguez-Boulan EJ: Membrane biogenesis, enveloped RNA 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 1403 viruses and epithelial polarity. In Modern Cell Biology, Vol 1, Satir BH, editor. New York, Alan R Liss, 1982, pp. 119-170. Rodriguez-Boulan E, Green RF, Meiss HK, and Sabatini DD: Enveloped viruses as tools for the study of epithelial polarity. In Perspectives in Differentiation and Hypertrophy, Anderson W and Sadler W, editors, Elsevier Science Publishing, 1982, pp. 51-64. Rodriguez-Boulan E, Paskiet KT, Salas PJI, and Bard E: Intracellular transport of influenza virus hemagglutinin to the apical surface of Madin-Darby canine kidney cells. J Cell Biol 98:308, 1984. Besharse JC: The daily light-dark cycle and rhythmic metabolism in the photoreceptor-pigment epithelium complex. In Progress in Retinal Research, Vol 1, Osborne N and Chader G, editors. New York, Pergamon Press, 1982, pp. 81-124. Besharse JC, Hollyfield JG, and Rayborn ME: Turnover of rod photoreceptor outer segments. II. Membrane addition and loss in relationship to light. J Cell Biol 75:507, 1977. Pfenninger KH and Maylie-Pfenninger MF: Lectin labeling of sprouting neurons. II. Relative movement and appearance of glycoconjugates during plasmalemmal expansion. J Cell Biol 89: 547, 1981. Hall MO, Bok D, and Bacharach ADE: Biosynthesis and assembly of the rod outer segment membrane system: formation and fate of visual pigment in the frog retina. J Mol Biol 45:397, 1969. Young RW and Droz B: The renewal of protein in retinal rods and cones. J Cell Biol 39:169, 1968. Young RW: Visual cells and the concept of renewal. Invest Ophthalmol 15:700, 1976. Young RW: Passage of newly formed protein through the connecting cilium of retinal rods in the frog. J Ultrastruct Res 23: 462, 1968. Papermaster DS, Converse CA, and Siu J: Membrane biosynthesis in the frog retina: opsin transport in the photoreceptor cell. Biochemistry 14:1343, 1975. Papermaster DS, Converse CA, and Zorn M: Biosynthetic and immunochemical characterization of a large protein in frog and cattle rod outer segment membranes. Exp Eye Res 23:105, 1976. Besharse JC and Pfenninger KH: Membrane assembly in retinal photoreceptors. I. Freeze-fracture analysis of cytoplasmic vesicles in relationship to disc assembly. J Cell Biol 87:451, 1980. Holtzman E, Schacher S, Evans J, and Teichberg S: Origin and fate of membranes of secretion granules and synaptic vesicles: membrane circulation in neurons, gland cells and retinal photoreceptors. In The Synthesis, Assembly and Turnover of Cell Surface Components, Poste G and Nicholson GL, editors. New York, Elsevier-North Holland, 1977, pp. 165-246. Kinney MS and Fisher SK: The photoreceptors and pigment epithelium of the larval Xenopus retina: morphogenesis and outer segment renewal. Proc R Soc Lond [Biol] 201:149, 1978. Nieuwkoop PD and Faber J: Normal Table of Xenopus laevis (Dandin). Amsterdam, North Holland Publishing, 1967, p. 245. Peters T Jr and Ashley CA: An artefact in radioautography due to binding of free amino acids to tissues by fixatives. J Cell Biol 33:53, 1967. Spurr AR: A low-viscosity epoxy resin embedding medium for electron microscopy. J Ultrastruct Res 26:31, 1969. Papermaster DS, Schneider BG, Zorn MA, and Kraehenbuhl JP: Immunocytochemical localization of opsin in the outer segments and Golgi zones of frog photoreceptor cells. J Cell Biol 77:196, 1978. Schneider BG and Papermaster DS: Immunocytochemistry of retinal membrane protein biosynthesis at the electron microscopic level by the albumin embedding technique. Methods Enzymol 96:485, 1983. Altman LA, Schneider BG, and Papermaster DS: Rapid embed- Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017 1404 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1985 ding of tissues in Lowicryl K.4M for immunoelectron microscopy. J Histochem Cytochem 32:1217, 1984. Carlemalm E, Garvito M, and Villiger W: Resin development for electron microscopy and an analysis of embedding at low temperature. J Microsc 126:123, 1982. Salpeter MM and Bachmann L: Autoradiography. In Principles and Techniques of Electron Microscopy: Biological Applications, Vol 2, Hayat MA, editor. New York, Van Nostrand Reinhold, 1972, pp. 219-278. Kopriwa BM: A comparison of various procedures for fine grain development in electron microscopic radioautography. Histochemistry 44:201, 1975. Salpeter MM, Bachmann L, and Salpeter EE: Resolution in electron microscope radioautography. J Cell Biol 41:1, 1969. Blackett NM and Parry DM: A new method for analyzing electron microscope autoradiographs using hypothetical grain distributions. J Cell Biol 57:9, 1973. Blackett NM and Parry DM: A simplified method of "hypothetical grain" analysis of electron microscope autoradiographs. J Histochem Cytochem 25:206, 1977. Salpeter MM, McHenry FA, and Salpeter EE: Resolution of electron microscope autoradiography. IV. Application to analysis of autoradiographs. J Cell Biol 76:127, 1978. Williams MA: The assessment of electron microscope autoradiographs. Adv Opt Elect Microsc 3:219, 1969. Papermaster DS and Schneider BG: Biosynthesis and morphogenesis of outer segment membranes in vertebrate photoreceptor cells. In Cell Biology of the Eye, McDevitt D, editor. New York, Academic Press, 1982, pp. 475-531. Papermaster DS, Reilly P, and Schneider BG: Cone lamellae and red and green rod outer segment disks contain a large intrinsic membrane protein on their margins: an ultrastructural immunocytochemical study of frog retinas. Vision Res 22:1417, 1982. Ainsworth SK and Karnovsky MJ: An ultrastructural staining method for enhancing the size and electron opacity of ferritin in thin sections. J Histochem Cytochem 20:225, 1972. Mercurio AM and Holtzman E: Ultrastructural localization of glycerolipid synthesis in rod cells of the isolated frog retina. J Neurocytol 11:295, 1982. Vol. 26 44. Young RW: The renewal of photoreceptor outer segments. J Cell Biol 33:61, 1967. 45. Chaitin MH, Schneider BG, Hall MO, and Papermaster DS: Actin in the photoreceptor connecting cilium: immunocytochemical localization to the site of outer segment disk formation. J Cell Biol 99:239, 1984. 46. Nir I and Papermaster DS: Differential distribution of opsin in the plasma membrane of frog photoreceptors: an immunocytochemical study. Invest Ophthalmol Vis Sci 24:868, 1983. 47. Peters KR, Palade GE, Schneider BG, and Papermaster DS: Fine structure of a periciliary ridge complex of frog retinal rod cells revealed by ultrahigh resolution scanning electron microscopy. J Cell Biol 96:265, 1983. 48. Defoe DM and Besharse JC: Membrane assembly in retinal photoreceptors. II. Immunocytochemical analysis of freeze-fractured rod photoreceptor membranes using anti-opsin antibodies. J Neurosci 1985, 5:1023. 49. Bok D, Hall MO, and O'Brien P: The biosynthesis of rhodopsin as studied by membrane renewal in rod outer segments. In International Cell Biology (1976-1977), Brinkley BR and Porter KR, editors. New York, Rockefeller University Press, 1977, pp. 608-617. 50. Fukuda MN, Papermaster DS, and Hargrave PA: Rhodopsin carbohydrate: structure of small oligosaccharides attached at two sites near the NH 2 -terminus. J Biol Chem 254:8201, 1979. 51. Liang CJ, Yamashita K, Muellenberg CG, Shichi H, and Kobata A: Structure of the carbohydrate moieties of bovine rhodopsin. J Biol Chem 254:6414, 1979. 52. Besharse JC and Forestner DM: Horseradish peroxidase uptake by rod photoreceptor inner segments accompanies outer segment disc assembly. 39th Ann Proc Electron Microsc Soc Am 486, 1981. 53. Holtzman E and Mercurio AM: Membrane circulation in neurons and photoreceptors: some unresolved issues. Int Rev Cytol 67: 1, 1980. 54. Besharse JC, Hollyfield JG, and Rayborn ME: Turnover of rod photoreceptor outer segments: accelerated membrane renewal in rods after exposure to light. Science 196:536, 1977. Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933115/ on 06/18/2017