Lectin receptors of rods and cones - IOVS

Lectin receptors of rods and cones Visualization by fluorescent label C.D.B. Bridges The binding of eight fluorescence-labeled lectins to visual cells was investigated. Concanavalin A (Con A) (specific for oligosaccharides containing niannose) bound evenly to the surfaces of rod outer segments (ROSs) from frog, cattle, goldfish, and turtle (Pseudemys). In frog, binding was observed on the inner segment, with intensifications at the inner/outer segment junction and in the region of the nucleus. No binding occurred on the surfaces ofROS discs. In 0.1% Triton X-100, rods formed convoluted tubular structures that bound Con A to their surfaces. Isolated cone outer segments (COSs) showed uniform fluorescence as compared with the inner segments, ivhich displayed only surface labeling. Only lectins with an affinity for fucose (UeA, LTA) failed to bind to the ROSs and COS. Lectins with an affinity for oligosaccharides containing galactose (RCA-120, RCA-60, and PNA) bound hardly at all to the rod inner segments. Both WGA (N-acetyl glucosamine) and SBA (N-acetyl galactosamine and galactose) bound indiscriminately to all ROSs, whereas PNA bound preferentially to the accessory cones. Key words: rods, cones, outer segments, inner segments, Mueller cells, lectins, fluorescence microscopy

T he glycoproteins and glycolipids of cell plasma membranes are oriented with their oligosaccharides extending outward into the extracellular space.1 These compounds are believed to participate in a variety of cell functions that include immunologic phenomena, recognition, development, contact inhibition, synaptic function, cancerous proliferation, and reactions to hormones and toxins. In the present work, the oligosaccharides on photoreceptor cell surfaces were investigated by means of fluorescence-laFrom the Cullen Eye Institute, Baylor College of Medicine, Department of Ophthalmology, Houston, Texas. This work was supported by grants from the Retina Research Foundation of Houston and by National Institutes of Health grants EY 02489 and EY 02520. Submitted for publication Dec. 17, 1979. Reprint requests: Dr. C. D. B. Bridges, Department of Ophthalmology, Baylor College of Medicine, Houston, Texas 77030.

8

beled sugar-specific lectins.2 Some lectins displayed uniform binding to the surfaces of rods and cones, whereas others bound differentially to rod outer segments (ROSs). Among cone types, the accessory cones were found to have a special affinity for peanut lectin. Materials and methods Fluorescein isothiocyanate (FITC) lectins were obtained from a variety of commercial sources: concanavalin A (Con A), Lotus agglutinin (LTA), wheat germ agglutinin (WGA), and soybean agglutinin (SBA) from Miles-Yeda; Ulex agglutinin specific for L-fucose (UeA), peanut agglutinin (PNA), and WGA from L'Industrie Biologique Francaise (L'lBF); 120,000 MW agglutinin from castor bean (RCA-120), 60,000 MW agglutinin from castor bean (RCA-60), and LTA from Sigma Chemical Co. Although the affinity of a lectin for its receptor may depend on extremely complex structural requirements, 3 a simplified summary of lectin specificities is given in Table I.

0146-0404/81/010008+09$00.90/0 © 1981 Assoc. for Res. in Vis. and Ophthal., Inc.

Volume 20 Number 1

Haptene sugar inhibitors used in the present work were: a-methyl mannopyranoside (200 mM in Ringer's solution; Calbiochem); disaccharide of N-acetyl glucosamine, chitobiose (10 mM; L'lBF); D-galactose (D-Gal) (100 mM; L'lBF); N-acetyl D-galactosamine (GalNAc) (100 mM; L'lBF); D-glucose (100 mM; Fisher); lactose (50 mM; L'lBF); a-D-fucose (a-D-fuc) and a-L-fucose (50 mM; Sigma); N-acetyl D-glucosamine (GlcNAc) (100 mM; L'lBF); a-methyl-D-galactopyranoside (Me-a-D-gal) (50 mM; Sigma); /3-methyl-D -galactopyranoside (Me-j3-D-gal) (50 mM; Sigma). Except where indicated in the Results, freshly dissected retinas were used. Frogs (Rana pipiens) were usually dark-adapted for 3 to 12 hr, and the retinas separated in dim red light under modified Ringer's solution.4 Cattle eyes were placed fresh on ice in the slaughterhouse, and the retinas were dissected within 2 to 3 hr. Either the whole retina was incubated in the FITC-lectin in Ringer's at concentrations ranging from 0.5 to 612 /u,g/ml (see Table II), or the ROSs were detached by quickly dipping the retina in and out of a drop of Ringer's on a glass slide. In the latter event, the resulting suspension was mixed with the FITC-lectin and coverslipped. Alternatively, the lectin was run in from the side after coverslipping. The latter technique avoided much of the agglutination observed with the first approach. 5 After incubation at room temperature for 5 to 15 min, the surplus lectin was rinsed out by leaving the coverslip in place and gently irrigating with fresh Ringer's until no background fluorescence was visible. Thorough rinsing prior to incubation with lectin was sometimes carried out without affecting the results. When required, haptene sugars were added with the lectin or to the rinsing solution. N,N'didansyl cystine (DDC) was used in 10 /xM concentration, 6 generally in combination with FITCCon A. Occasionally rods were isolated by shaking the retina in a solution of 1% glutaraldehyde dissolved in Ringer's at room temperature. Many rods detached at a point just below the nucleus. After several washings with Ringer's, the organelles were then incubated with lectin as described above. Binding of the lectin occurred in the normal way in these preparations and was inhibited by the appropriate haptene sugars. Fluorescence was observed with a Zeiss Universal microscope equipped for epifluorescence. Illumination was provided by xenon or mercury lamps. Various dichroic/filter combinations were used to excite and isolate FITC and DDC fluores-

Lectin receptors of rods and cones

9

Table I. Summary of lectin affinities* Lectin Con A

WGA SBA PNA RCA-120 and RCA-60 HPA DBA LTA UeA

Affinity for a-Man, a-Glc, disaccharides and trisaccharides of a(l —• 2) Man; high affinity for terminal GlcNAc linked /3(1 -> 2) to Man j8(l -+ 4)-linked oligomers of GlcNAc Terminal a- or 0-GalNAc or Gal Terminal a-Gal and Gal-/3 (1 -> 3) GalNAc Terminal )3-Gal. GalNAc inhibits binding of RCA-60 to its receptors, but not RCA-120 Terminal a-GalNAc Terminal a-GalNAc Terminal a-L-fuc, /3-L-fuc Terminal a-L-fuc, /3-L-fuc

*Helix pomatia and Dolichos biflorus agglutinins were not used in the present work (see ref. 5).

cence. Black and white photography was carried out with Kodak Tri-X film pushed to an effective exposure index of 1600 by developing in Diafine (Acufine, Inc., Chicago, 111.). Exposure times ranged from 5 to 45 sec, averaging about 15 to 25. Little improvement was obtained beyond 30 sec, owing to fading of the FITC chromophore. In addition, fresh rods disintegrated during prolonged exposure to the irradiating light (440 to 490 nm). Results Con A Rods—fresh and fixed. Fig. 1 is a surface view of a fresh frog dark-adapted retina incubated with FITC-Con A. The fluorescence was distributed as a bright line defining the surface of the ROS, showing that the lectin had bound to the plasma membrane. Labeling did not occur if 100 mM a-methyl mannopyranoside was present. Fig. 2 is the edge of a fresh retina incubated with FITC-Con A. Because this material was light-adapted, part of the fluorescence was obscured by melanin granules in the processes torn from the pigment epithelium during isolation of the retina. The surface labeling of the ROSs was again visible. The inner bright band (arrow) indicates the level of the external limiting membrane. As in Fig. 1, binding was inhibited by 100 mM

10

Invest. Ophthalmol. Vis. Sri. January 1981

Bridges

Figs. 1 through 4. For legend see facing page.

Volume 20 Number 1


11

Table II. Fluorescence of photoreceptors i ncubated with FITC-labeled lectinsA Rods Lectin

Concentration (fig/ml)B

Con A WGA SBA PNA UeAG RCA-120 LTA RCA-60

0.5-200 100-250 63-200 5-200 612 25-100 250 25-100

OS

Cones IS

OS

IS = inner segment. A Fluorescence intensities were graded arbitrarily on the basis of visual judgment. + + + + + is the strongest, and + is marginally detectable. B Note that all concentrations are expressed in micrograms per milliliter. This is considered preferable to converting them to molar concentrations, since although the molecular weights of many lectins are known (Con A, 102,000; HPA, 79,000; Ulex, 170,000; RCA-120, 120,000; RCA-60, 60,000; WGA, 36,000; SBA, 120,000; and PNA, 110,000), others have not been firmly established. An example is Dolichos biflorus agglutinin (110,000 or 135,000). Further, the LTA used here consisted of a mixture of three fueose-specific lectins of molecular weight, respectively, 120,000, 58,000, and 117,000. An added complication is that lectins are multivalent—WGA has two binding sites, whereas Helix pomatia agglutinin has six. c I n rod cells isolated from papain-dissociated retinas, Con A was taken up evenly on the surface down to and including the synaptic pedicle. D Evenly fluorescent in detached principal and accessory cones; fluorescence less penetrant and more confined to the surfaces of cones still attached to the retina or fixed in glutaraldehyde (this also applied to goldfish and Pseudemys cones). E Accessory cone: principal cone much weaker. ••"Fluorescence often appeared patchy. G Brightly fluorescent, amorphous material was visible in many of these preparations (similar material was seen on the retinal surface of the retinal pigment epithelium). "Principal and accessory cones. 'IS also.

a-methyl mannopyranoside but not by any of the other sugars tested, except for slight inhibition by 700 mM glucose. Isolated rods were sometimes obtained by shaking the retina in a solution of 1% glutaraldehyde in Ringer's. Fig. 3 illustrates one such rod that has been incubated in FITCCon A. The bright region dividing inner and outer segments was also visible in fresh rods7 but was obliterated in papain-dissociated photoreceptors. In this material, Con A binding was visible as far as the synaptic pedicle. In Fig. 3 there are about eight lines running parallel to the long axis of the outer segment. They were visible on the surface of fresh and

fixed ROS and ran for more than one third of their length. They probably represent the calycal processes that project from the inner segment and are either the calyces themselves or material in the grooves in which they run. Above the nucleus was a further intensification of fluorescence corresponding to the level of the external limiting membrane. Fig. 4 shows an isolated fresh bovine ROS. Although much narrower, its general appearance was like that of frog ROSs. Goldfish and Pseudemys ROSs bound FITC-Con A in a similar way. Binding to the bimembranous interior

Fig. 1. Fluorescence photographs of the surface of a dark-adapted frog retina (R. pipiens) after incubation at room temperature with FITC-Con A (500 /ng/ml for 3 min). The black flecks visible in b are melanin granules. (Both bars = 30 /u-m.) Fig. 2. Edge view of light-adapted frog retina treated as in Fig. 1. (Bar = 50 /xm.) Fig. 3. Frog rod isolated by shaking retina in 1% glutaraldehyde and incubated in FITC-Con A, 100 /u-g/ml. O, ROS; I, inner segment; N, nucleus. (Bar = 10 jam.) Fig. 4. Bovine ROS incubated in FITC-Con A, 100 /xg/ml. (Bar = 10 fxm.)

12

Bridges

Invest. Ophthalmol. Vis. Sci. January 1981

Fig. 5. Frog ROS after exposure to 10% Ringer's containing Con A (100 /tig/ml) and DDC (10 JJLM). a, transmitted light (phase), b, FITC fluorescence, c, DDC fluorescence. (Bar = 25 /u.m.) Fig. 6. Effects of detergent. Frog ROS incubated in Ringer's containing 0.1% Triton X-100 and 50 ju.g/ml FITC-Con A. a and c, Transmitted light (phase), b and d, FITC fluorescence. (Bar = 25 /u.m.) Fig. 7. Isolated frog cones incubated in FITC-Con A, 50 fig/m\. (Bar = 15 fxm.) Fig. 8. Surface view of frog retina incubated with FITC-WGA (500 ^ig/ml) showing uptake of label by the COSs. (Bar = 20 pm.)

Volume 20 Number 1


Fig. 9. Surface view of frog retina incubated with FITC-PNA (200 /xg/ml). (Bar = 50 /im.) Fig. 10. Isolated frog ROS labeled with FITC-PNA as in Fig. 9. Slight patching is visible, with some intensification of fluorescence at one end. (Bar = 10 pirn.) Fig. 11. External limiting membrane viewed from the retinal surface after incubation with FITC-RCA-120 (100 Aig/ml). The relatively faint labeling of the out-of-focus pbotoreceptors renders them invisible against the intense fluorescence from the Mueller cell ciliary processes. (Bar = 20 fim.) Fig. 12. Edge view of frog retina incubated with FITC-PNA (200 jag/ml). The bright, unfocused objects in the lower sector are COSs. (Bar = 50 fim.) Fig. 13. Frog COSs labeled with FITC-PNA (200 /ag/ml). This is a surface view of the retina where the photoreceptors have been deflected to one side by slight pressure from the coverslip. The relatively lightly labeled ROSs are invisible under these conditions. (Bar = 35 /im.)

Fig. 14. Frog COSs labeled with FITC-PNA (200 /Ag/ml). A transmitted light, phase-contrast exposure was taken on the same frame as the fluorescence photograph, so visualizing the oil droplets in the principal and single cones. Since no fluorescent COS appears to be aligned with an oil droplet, these organelles belong to the accessory cones, (Bar = 20 /xm.)

13

14

Bridges

discs was not observed. Fig. 5 illustrates a rod that has been exposed to a strongly hypotonic medium in the presence of FITCCon A mixed with the fluorescent probe DDC.6 Fig. 5, a, was taken in normal phase; Fig. 5, b, in the fluorescent light emitted by FITC; and Fig. 5, c, in the light emitted by DDC. Fig. 5, a and c, shows the extent of the membranous material present, but binding of Con A was localized on what appear to be the torn and folded remnants of the plasma membrane (Fig. 5, b). After incubation with Con A, the surfaces of ROSs tended to slough off. When this occurred, the interior stacked discs freely took up DDC, but they did not do so when the surface Con A binding was even and unbroken.8 Completely stripped ROSs retained their shape, indicating the existence of an interdiscal "adhesive. "9 Fig. 6 illustrates the action of low concentrations of detergents on ROSs. In the presence of 0.1% Triton X-100, the rods formed tubular structures that became highly convoluted. If FITC-Con A was present, the tubes carried a strong surface fluorescence (Fig. 6, b), but their genesis from the bimembranous discs of the original ROS has not been clarified. Later, the Con A label became patchy and disjointed, as seen in Fig. 6, d. Cones. Fig. 7 depicts four freshly isolated cones incubated with FITC-Con A. Unlike that of rods, the COS fluorescence was uniform, not confined to the plasma membrane as in the cone inner segments. The labeling was inhibited by 100 mM a-methyl mannopyranoside. Other lectins. WGA, SBA, PNA, UeA, LTA, RCA-120, and RCA-60 were investigated in addition to Con A. Neither UeA nor LTA bound to rods and cones (Table II). However, the other five lectins bound to various degrees. SBA bound weakly to ROSs but had a stronger affinity for COSs. WGA bound to rod outer and inner segments. This lectin also had a high affinity for COSs, as shown in Fig. 8. In this figure the COS pairs represent principal/accessory cone combinations. PNA, RCA-120, and RCA-60 (specific for


galactose) bound to ROS. Fig. 9 is the surface view of a retina incubated with FITC-PNA. Its appearance was comparable with the retina incubated with Con A in Fig. 1, but the fluorescence intensity was much weaker (see also Table II). As shown in Fig. 10, binding of PNA was intensified at one end of the ROSs (the proximal end) and was sometimes patchy, perhaps indicating aggregation of the lectin receptors. As noted previously,7 RCA-120 and PNA differed from the other lectins tested in binding preferentially to the ROS plasma membranes. The rod inner segments displayed little surface fluorescence. Similar results were obtained with RCA-60 in the present study. PNA and the ricins bound strongly to the external limiting membrane. Fig. 11 shows the appearance of the external limiting membrane after incubation with RCA-120. The invisible photoreceptors passed through the black holes in the intensely fluorescent reticulum formed from the terminal processes of the Mueller cells. In Mueller cells isolated from dissociated retinas, these processes have been shown to have a strong affinity for a variety of lectins.I0 PNA has a strong affinity for the accessory cones. Fig. 12 is a semi-edge fold of a fresh frog retina incubated with PNA. The accessory cones appeared as intensely fluorescent structures dotted over the preparation. These were clearly visible in the surface view of Fig. 13, where the photoreceptors were flattened slightly by pressure from the coverslip. Fig. 14 is a double exposure taken first in fluorescence, then under low-intensity transmitted light. Although cone inner segment oil droplets were abundant over the whole field, none were aligned with a fluorescent COS. The COSs with oil droplets— the principal and single cones—therefore had little affinity for PNA. As with Con A, the effects of various haptene sugars were tested: binding of WGA was inhibited by the GlcNAc disaccharide chitobiose (10 mM) and GlcNAc (100 mM); SBA by GalNAc (100 mM); PNA, RCA-120, and

Volume 20 Number 1

RCA-60 by D-Gal (100 mM), Me-a-D-gal (50 mM), Me-/3-D-gal (50 mM), a-D-fuc (50 mM), and lactose (50 mM). Discussion These findings indicate that receptors for a wide variety of lectins exist on the surfaces of rods and cones. Although the abundance of these receptors cannot be judged from subjective observations of fluorescence, a rough ordering of fluorescence intensities is indicated in Table II. The number of FITC residues attached to each lectin molecule ranged from 1.6 for Con A to about 14 for WGA (suppliers' specifications). However, the strongest fluorescence was with Con A, which had the lowest specific label, and the weakest fluorescence was with SBA, which had a higher specific label (about 8). As reported elsewhere, 7 ' u the surfaces of ROSs label more intensely than the inner segments when incubated with FITC-RCA120 or PNA. The same effect was observed here with RCA-60. These three lectins share an affinity for oligosaccharides containing terminal Gal and/or GalNAc (Table I), which suggests that these sugars are either absent from the inner segment plasma membrane or are inaccessible. The difference between inner and outer segment membranes suggests that membrane receptors containing accessible Gal and GalNAc are not free to flow from outer to inner segment. 7 Because of this restriction and because many different glycoproteins have Man-GlcNAc residues, 12 the Con A receptors on the inner segment plasma membrane may differ from those in the outer segment. The Con A and WGA receptor in the ROS may be rhodopsin, 13 which has the requisite combination of GlcNAc and Man residues.14- 15 The 291,000-dalton protein16- 17 also binds Con A and WGA13 but is present in only minor proportion. Rod and cone discs are formed by infolding of the plasma membrane. In rods these infoldings are sealed off to form stacks of flattened bimembranous discs, so that the oligosaccharide layer that normally resides on the extracellular surface18 is sequestered within


15

the disc interiors. Thus, in the experiment of Fig. 5, Con A did not bind to the disc surfaces, confirming Rohlich's observations.19 In cones the discs are confluent with the extracellular space. This permits penetration of FITC-lectins into their interiors and accounts for their uniformly fluorescent appearance. The fluorescence is frequently more marked in isolated cells, suggesting that after detachment the intradiscal space expands (if kept for several hours, isolated COSs appear to "unravel"). Since Con A and WGA bind to cones, their visual pigments may contain oligosaccharide sequences similar to those found in rhodopsin, but perhaps with different lectin affinities. Iodopsin and rhodopsin, for example, are eluted from columns of immobilized Con A by different concentrations of a-methyl mannopyranoside. 20 Since the dimensions of accessory cone membranes are indistinguishable from those of principal and single cones, the preferential binding of PNA, which contrasts with WGA, suggests an excess of Gal-/3(1 —» 3) GalNAc residues on the accessory cones compared with a more equal distribution of GlcNAc-/3 (1 —» 4) GlcNAc between the two receptor classes. The fucose-specific lectins UeA and LTA did not bind to rods or cones, although both lectins bound extensively to the retinal surface of the pigment epithelial cells and to amorphous material that adhered to this surface. Sheets and occasional chunks of this matter were sometimes found on the detached retina and probably represented the interphotoreceptor matrix, which is known to contain fucose residues. 21 REFERENCES 1. Harmon RE: Cell Surface Carbohydrate Chemistry. New York, 1978, Academic Press, Inc. 2. Goldstein IJ and Hayes CE: The lectins: carbohydrate-binding proteins of plants and animals. Adv Carbohydr Chem Biochem 35:127, 1978. 3. Dulaney JT: Binding interactions of glycoproteins with lectins. Mol Cell Biochem 21:43, 1978. 4. Bridges CDB: Vitamin A and the role of the pigment epithelium during bleaching and regeneration of rhodopsin in the frog eye. Exp Eye Res 22:435, 1976.

16


Bridges

5. Bridges CDB: Agglutination of isolated rod outer segments by lectins. INVEST OPHTHALMOL VIS SCI

20:17, 1981. 6. Yoshikami S, Robinson WE, and Hagins WA: Topology of the outer segment membranes of retinal rods and cones revealed by a fluorescent probe. Science 185:1176, 1974. 7. Bridges CDB and Fong S-L: Different distribution of receptors for peanut and ricin agglutinins between inner and outer segments of rod cells. Nature 282:513, 1979. 8. Bridges CDB: Reactions of isolated rods and pigment epithelium cells with fluorescence-labeled

14.

15.

16.

concanavalin A. INVEST OPHTHALMOL VIS SCI 17

(ARVO Suppl.):156, 1978. 9. Cohen AI: Chemo-surgical studies on outer segments. In, Biochemistry and Physiology of Visual Pigments, Langer H, editor. New York, 1973, Springer-Verlag, pp. 285-294. 10. Sarthy PV, Bridges CDB, Lam DMK, and Kretzer F: Lectin receptors on cells isolated from turtle ret-

17.

18.

inas. INVEST OPHTHALMOL Vis SCI 18(ARVO Suppl.):

259, 1979. 11. Nir I and Hall MO: Ultrastructural localization of lectin binding sites on the surface of retinal photoreceptors and pigment epithelium. Exp Eye Res 29:181, 1979. 12. Montreuil J: Recent data on the structure of the carbohydrate moiety of glycoproteins. Metabolic and biological implications. Pure Appl Chem 42: 431, 1975. 13. Bridges CDB and Fong S-L: Lectins as probes of

19.

20.

21.

glycoprotein and glycolipid oligosaccharides in rods and cones. Neurocheinistryl:255, 1980. Liang CJ, Yamashita K, Shichi H, Muellenberg CG, and Kobata A: Structure of the carbohydrate moieties of bovine rhodopsin. J Biol Chem 254:6414, 1979. Fukuda MN, Papennaster 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. Bownds D, Brodie A, Robinson WE, Plainer D, Miller J, and Shedlovsky A: Physiology and enzymology of frog photoreceptor membranes. Exp Eye Res 18:153, 1974. Dreyer WJ, Papermaster DS, and Kuhn H: On the absence of ubiquitous structural protein subunits in biological membranes. Ann NY Acad Sci 195:61, 1972. Hirano H, Parkhouse B, Nicolson GL, Lennos ES, and Singer SJ: Distribution of saccharide residues on membrane fragments from a myeloma cell homogenate: its implications for membrane biogenesis. Proc Natl Acad Sci USA 69:2945, 1972. Rohlich P: Photoreceptor membrane carbohydrate on the intradiscal surface of retinal rod disks. Nature 263:789, 1976. Fager LY and Fager RS: Separation of rod and cone pigments from the chicken retina. INVEST OPHTHALMOL VIS SCI 17(ARVO Suppl.):126, 1978. Feeney L: Synthesis of interphotoreceptor matrix. I. Autoradiography of3H-fi.icose incorporation. INVEST OPHTHALMOL 12:739, 1973.