Molecular Vision 1998; 4:28 <http://www.molvis.org/molvis/v4/p28> Received 22 May 1998 | Accepted 5 December 1998 | Published 10 December 1998

Download Reprint

Glycosylation and Palmitoylation Are Not Required for the Formation of the X-Linked Cone Opsin Visual Pigments

Harry Ostrer,¹ Raju K. Pullarkat,² Manija A. Kazmi¹
 
 

¹Human Genetics Program, Department of Pediatrics, New York University Medical Center, New York, NY; ²Department of Neurochemistry, Institute for Basic Research in Developmental Disabilities, Staten Island, NY

Correspondence to: Harry Ostrer, MD, Human Genetics Program, Department of Pediatrics, New York University School of Medicine, 550 First Avenue - MSB136, New York, NY 10016; Phone: (212) 263-5746; FAX: (212) 263-7590; email: ostreh01@mcrcr6.med.nyu.edu

Abstract

Purpose: This study was designed to test whether palmitoylation and glycosylation are required for the formation of the green opsin visual pigment.

Methods: Stable cell lines were established by transfecting EBNA-293 cells with a pMEP4ß recombinant plasmid containing wild-type bovine rhodopsin or wild-type or mutant (N32S) green opsin cDNA molecules that included a tag for the eight amino acid residues located at the C-terminus of rhodopsin. The opsins were induced by addition of CdCl₂ into the medium and then reconstituted with 11-cis-retinal. The reconstituted opsins were purified by immunoaffinity chromatography, then analyzed by difference spectra, and by binding ³⁵S-GTP in the presence of bovine transducin. Non-reconstituted opsins were analyzed by western blotting and by pulse-labeling with ³H-palmitic acid followed by immunoprecipitation.

Results: Elimination of glycosylation by mutagenesis of the N-linked glycosylation site did not impair the ability of the resulting cone opsin to absorb light at the appropriate wavelength nor to activate transducin. Furthermore, as judged by pulse-labeling with ³H-palmitic acid and immunoprecipitation and by gas chromatography-mass spectroscopy, the wild type green opsin differs from rhodopsin by not being palmitoylated.

Conclusions: Glycosylation and palmitoylation are not required for the formation of cone opsin visual pigments. For the previously described green opsin C203R mutation, disruption of folding and transport, rather than altered glycosylation is sufficient to explain the associated color vision deficiency.

Introduction

The visual pigments are a family of photoreceptor proteins that absorb light and mediate vision [1-7]. Humans have two groups of visual pigments, rhodopsin, which is expressed in rod cells and provides monochromatic vision under low intensity light, and the cone opsins, which provide color vision under higher intensity light. There are three cone opsin pigments with short (S cone or blue), medium (M cone or green), or long (L cone or red) wavelength absorption spectra, each encoded by a separate gene [8]. The genes for red and green cone opsins are found in a tandem array on the distal long arm of the X chromosome, which may contain varying copy numbers of red, green and red/green hybrid genes; the blue opsin gene is a single-copy locus located at 7q22-qter [8-10]. All of the visual pigments are composed of an apoprotein molecule (or opsin) that is conjugated to the chromophore, 11-cis-retinal. In response to the absorption of a photon of light, the chromophore is isomerized to all-trans-retinal. This conformational change in the visual pigment molecule causes activation of the G-binding protein, transducin.

The structure and function of the rhodopsin molecule have been extensively studied by peptide mapping and site-directed mutagenesis and is predicted to be a protein with seven transmembrane segments. The cone opsins share 40-44% homology with rhodopsin at the amino acid level, suggesting similar structures (The structure for the green opsin gene is shown in Figure 1) [8]. Regions of structural similarity between the X-linked cone opsins and rhodopsin include a lysine at position 312, which is predicted to form a Schiff base with retinal, and cysteines at residues 126 and 203 that function as sites of disulfide cross-linking [11-15]. The comparable sites in rhodopsin are a lysine at position 296 and cysteines at residues 110 and 187 [12].

Previously it has been shown that both rhodopsin and the X-linked cone opsins are N-linked glycosylated and that rhodopsin is palmitoylated [16-20]. Rhodopsin is glycosylated at residues N2 and N15, although glycosylation only at asparagine-15 is required to form a visual pigment. Here, we show that elimination of the N-linked glycosylation does not impair the ability of green opsin to form a visual pigment and that the wild-type green opsin is not palmitoylated.

Methods

Plasmids

Recombinant plasmids included a full-length, synthetic bovine rhodopsin cDNA (pRho), a mutant bovine rhodopsin cDNA (pRhoC187R), and a full-length green cDNA (pGrn) that included the epitope for the C-terminus of rhodopsin [15,16,21]. Alteration of the N-linked glycosylation site from asparagine to serine at residue 32 (N32S, see Figure 1) was performed by site-directed mutagenesis (Transformer Site-Directed Mutagenesis Kit, Clontech Laboratories, Palo Alto, CA). The selection primer for transforming a MluI site to a HindIII site was 5'-CGACGGTATCGATAAGCTTGATATCGAATTCC-3'.

The primer for mutagenesis (N32S) was 5'-CTCTGGTGGACTGGCTGTTGGTG-3'. All mutants were verified by sequencing. Recombinant cDNAs were subcloned into the vector, pMEP4ß (provided by Dr. Mark Tyckocinski, Case Western Reserve University, Cleveland, OH), for creation of stable cell lines [16,22,23]. The final constructs was called pGrn-N32S.

Transfection, cell culture, selection, induction, western blots, glycosidase digestion, spectra, transducin activation, and immunoprecipitation were performed as described previously [15,16]. To inhibit glycosylation, EBNA-293 cells stably transfected with pGrn or pGrn-N32S were treated with 0.5 µg/ml tunicamycin (Sigma-Aldrich, St. Louis, MO) during the 16-hour induction with CdCl₂. To test for palmitoylation, EBNA-293 cells stably transfected with the rhodopsin or cone opsin plasmids were induced with 5 µM CdCl₂, then labeled with in the presence of 0.1 mCi ³H-palmitic acid (New England Nuclear, Beverly, MA). Induction and labeling were for 16 hours.

Fatty acid extraction and analysis were based on a previously described method [19]. The fatty acid methyl esters were extracted with hexane and an aliquot was analyzed on a Hewlett-Packard 5890 gas chromatograph equipped with a flame ionization detector and a fused silica capillary column (0.25 mm x 30 m, coated with BD225 at a film thickness of 0.25 µm, J & W Scientific, Folsom, CA). The initial column temperature was 182 °C and programmed at 2 °C/min to a final temperature at 220 °C. Under the condition, methyl palmitate eluted at 8.95 min and methyl eicosanoic (an internal standard) eluted at 22.06 min. The quantity of palmitic acid was normalized to the initial amount of protein from which it was extracted [19].

Results

Glycosylation is not required for formation of a green opsin visual pigment

The epitope-tagged, wild-type and N32S green opsins were expressed in 293-EBNA cells. As judged by western blot analysis, the N32S mutant was not glycosylated (Figure 2A). Treatment with tunicamycin, which inhibits N-linked glycosylation, collapsed the pattern of the wild type green opsin and did not affect the pattern of the N32S mutant (Figure 2B). The photobleaching difference spectrum revealed a maximal absorption of light ([lambda]_max value) at 530 nm for both the wild type and N32S opsins, demonstrating that the unglycosylated cone opsin forms a visual pigment that is comparable to the wild type (Figure 3). The increased noise observed in the N32S mutant tracing suggests the possibility that the unglycosylated form is a less efficient pigment, perhaps due to less efficient folding. However, the function of the glycosylated and unglycosylated visual pigments were tested further by the GTP[gamma] (³⁵S) binding assay, which showed that both the wild-type and N32S green opsins activated bovine transducin to a comparable degree in response to light (Figure 4). Using least squares regression analysis, the slopes (± standard errors) are: Grn-WT 296.7±19.3, Grn-N32S 301.4±35.9, pMEP4ß 12.3±1.9. The test statistic (two sample t-test) for the difference in slopes between Grn-WT and Grn-N32S was 0.12 (p=0.91). The test statistic for the difference in slopes between Grn-WT and pMEP4ß was 14.9 (p<0.0001) and between Grn-N32S and pMEP4ß was 8.0 (p<0.001).

The green opsin visual pigment is not palmitoylated

To determine whether the green opsin visual pigment was palmitoylated, the epitope-tagged, wild-type green opsin, rhodopsin and mutant (C187R) rhodopsin were expressed in 293-EBNA cells in the presence of ³H-palmitic acid, immunoprecipitated using the 1D4 antibody and analyzed by polyacrylamide gel electrophoresis. A band was observed for rhodopsin, but not the green cone opsin, nor for the rhodopsin mutant C187R, a polypeptide that is known to be aberrantly folded and not palmitoylated (Figure 5). These findings were confirmed by direct chemical analysis of the fatty acids extracted from wild type rhodopsin and green opsin (Table 1, p =.035).

Discussion

These studies demonstrate that glycosylation is not required for the X-linked cone opsins to form a visual pigment. Because the green cone opsin is a membrane-bound protein, glycosylation of the protein occurs at its N-linked site during its passage through the endoplasmic reticulum, regardless of any functional requirements. Inhibition of rhodopsin glycosylation by treatment with tunicamycin did not affect in vivo folding, assembly, or transport of the opsin in COS-1 cells, nor the formation of a visual pigment with 11-cis-retinal; however, the unglycosylated rhodopsin was only 10% as efficient at achieving light-dependent activation of transducin at comparable concentration levels of the visual pigment. Of the two sites at which N-linked glycosylation occurs (N2 and N15), only N15 glycosylation appeared to be important for maximally efficient activation of transducin [17]. The absence of glycosylation at this site did not affect the formation of the MII species of rhodopsin, nor the stability of the retinyl-opsin linkage. Rather the lower stability of the unglycosylated MII intermediate compared to the glycosylated form appeared to account for these differences in transducin activation [17]. Likewise, unglycosylation of the ß₂-adrenergic receptor impaired agonist-dependent signal transduction [24]. The green cone opsin differs from these other 7-transmembranous segment receptor proteins by not requiring glycosylation for signal transduction; however, the possibility cannot be excluded that unglycosylation of the cone opsin might have subtle effects on the stability of the protein that were not tested in this series of experiments.

These studies also demonstrate that, unlike rhodopsin, the X-linked cone opsins are not palmitoylated. Palmitoylation of polypeptides usually occurs at the consensus site of paired cysteine residues, sites that are present in rhodopsin (C322 and C323) and absent in the cone opsins. However, for some proteins, such as the ß₂-adrenergic receptor, palmitoylation occurs at a single cysteine [25]. Although a cysteine is found at residue 332 in the cone opsin molecule, this amino acid is apparently not used for palmitoylation.

For rhodopsin, the paired palmitic acid residues are inserted in the plasma membrane, anchoring the C-terminal tail and creating a fourth cytoplasmic loop [19,20]. The absence of the cysteines suggests that the cone opsin does not have a comparable transmembranous region and thus is likely to have a different structure in its cytoplasmic C-terminal region. Differences in the palmitoylation and structure of the opsins could affect their physical properties. Receptor palmitoylation has been linked to desensitization, a process that includes phosphorylation and binding arrestin, an inhibitory protein, following exposure to the agonist or stimulus [26]. ß₂-adrenergic receptors in which the palmitoylation site has been mutated are more phosphorylated and less active in signaling than wild-type. Continued agonist exposure causes uncoupling of the wild type, but not the mutant, receptor to the G-binding protein. These findings suggest that the unpalmitoylated mutant protein is desensitized compared to the wild type [25].

A comparable link between photoreceptor desensitization and palmitoylation has not been demonstrated for rhodopsin. Rhodopsin mutants that were not palmitoylated did not to show either a decrease in transducin activation nor an increase in light-dependent phosphorylation [19]. By contrast, removal of palmitate from rhodopsin in the rod outer segment membranes by treatment with hydroxylamine led to an increase in light-dependent transducin binding, despite a reduction in visual pigment regeneration with the chromophore [27]. None of these studies demonstrates a link between photoreceptor desensitization and palmitoylation; however, the role of palmitoylation has not been tested for the other important aspect of photoreceptor desensitization, arrestin binding. Hence, the significance of palmitoylation for rhodopsin function remains undetermined.

Previously we have shown that the common mutation associated with color vision deficiencies, C203R, is unglycosylated, aberrantly folded and deficient in transport to the cell membrane [15]. The absence of glycosylation of this protein appears to be a consequence of the aberrant folding. Because glycosylation appears to have no effects on cone opsin visual pigment formation, the aberrant folding and transport are sufficient to explain the association of this mutation with color vision deficiencies [13,14].

Acknowledgements

The authors thank M. Han and T. Sakmar for advice with assays, C. Oddoux for helpful discussions, R. Pergolizzi for assistance with the automated DNA sequence analysis, R. Shore for assistance with the statistical analysis, and Mark Tyckocinski for providing the PMEP4ß vector.

References

1. Applebury ML, Hargrave PA. Molecular biology of the visual pigments. Vision Res 1986; 26:1881-95.

2. Nathans J. Molecular biology of visual pigments. Annu Rev Neurosci 1987; 10:163-94.

3. Nathans J. Protein-chromophore interactions in rhodopsin studied by site-directed mutagenesis. Cold Spring Harb Symp Quant Biol 1990; 55:621-33.

4. Hargrave PA, McDowell JH. Rhodopsin and phototransduction: a model system for G protein-linked receptors. FASEB J 1992; 6:2323-31.

5. Stryer L. Visual excitation and recovery. J Biol Chem 1991; 266:10711-4.

6. Khorana HG. Rhodopsin, photoreceptor of the rod cell. An emerging pattern for structure and function. J Biol Chem 1992; 267:1-4.

7. Nathans J. In the eye of the beholder: visual pigments and inherited variation in human vision. Cell 1994; 78:357-60.

8. Nathans J, Thomas D, Hogness DS. Molecular genetics of human color vision: the genes encoding blue, green, and red pigments. Science 1986; 232:193-202.

9. Nathans J, Piantanida TP, Eddy RL, Shows TB, Hogness DS. Molecular genetics of inherited variation in human color vision. Science 1986; 232:203-10.

10. Vollrath D, Nathans J, Davis RW. Tandem array of human visual pigment genes at Xq28. Science 1988; 240:1669-72.

11. Davidson FF, Loewen PC, Khorana HG. Structure and function in rhodopsin: replacement by alanine of cysteine residues 110 and 187, components of a conserved disulfide bond in rhodopsin, affects the light-activated metarhodopsin II state. Proc Natl Acad Sci U S A 1994; 91:4029-33.

12. Karnik SS, Khorana HG. Assembly of functional rhodopsin requires a disulfide bond between cysteine residues 110 and 187. J Biol Chem 1990; 265:17520-4.

13. Winderickx J, Sanocki E, Lindsey DT, Teller DY, Motulsky AG, Deeb SS. Defective colour vision associated with a missense mutation in the human green visual pigment gene. Nat Genet 1992; 1:251-6.

14.Nathans J, Maumenee IH, Zrenner E, Sadowski B, Sharpe LT, Lewis RA, Hansen E, Rosenberg T, Schwartz M, Heckenlively JR, et al. Genetic heterogeneity among blue-cone monochromats. Am J Hum Genet 1993; 53:987-1000.

15. Kazmi MA, Sakmar TP, Ostrer H. Mutation in a conserved cysteine in the X-linked cone opsins causes color vision deficiencies by disrupting protein folding and stability. Invest Ophthalmol Vis Sci 1997; 38:1074-81.

16. Kazmi MA, Dubin RA, Oddoux C, Ostrer H. High-level inducible expression of visual pigments in transfected cells. Biotechniques 1996; 21:304-11.

17. Kaushal S, Ridge KD, Khorana HG. Structure and function in rhodopsin: the role of asparagine-linked glycosylation. Proc Natl Acad Sci U S A 1994; 91:4024-8.

18. O'Brien PJ, Zatz M. Acylation of bovine rhodopsin by [3H] palmitic acid. J Biol Chem 1984; 259:5054-7

19. Karnik SS, Ridge KD, Bhattacharya S, Khorana HG. Palmitoylation of bovine opsin and its cysteine mutants in COS cells. Proc Natl Acad Sci U S A 1993; 90:40-4.

20. Ovchinnikov YuA, Abdulaev NG, Bogachuk AS. Two adjacent cysteine residues in the C-terminal cytoplasmic fragment of bovine rhodopsin are palmitylated. FEBS Lett 1988; 230:1-5.

21. Ostrer H, Kazmi MA. Mutation of a conserved proline disrupts the retinal-binding pocket of the X-linked cone opsins. Mol Vis 1997; 3:16 <http://www.emory.edu/molvis/v3/p16/>.

22. Hambor JE, Hauer CA, Shu HK, Groger RK, Kaplan DR, Tyckocinski ML. Use of an Epstein-Barr virus episomal replicon for anti-sense RNA-mediated gene inhibition in a human cytotoxic T-cell clone. Proc Natl Acad Sci U S A 1988; 85:4010-4.

23. Hauer CA, Getty RR, Tykocinski ML. Epstein-Barr virus episome-based promoter function in human myeloid cells. Nucleic Acids Res 1989; 17:1989-2003.

24. Boege F, Ward M, Jurss R, Hekman M, Helmreich EJ. Role of glycosylation for beta 2-adrenergic receptor function in A431 cells. J Biol Chem 1988; 263:9040-9.

25. O'Dowd BF, Hnatowich M, Caron MG, Lefkowitz RJ, Bouvier M. Palmitoylation of the human beta 2-adrenergic receptor. Mutation of Cys341 in the carboxyl tail leads to an uncoupled nonpalmitoylated form of the receptor. J Biol Chem 1989; 264:7564-9.

26. Mouillac B, Caron M, Bonin H, Dennis M, Bouvier M. Agonist-modulated palmitoylation of beta 2-adrenergic receptor in Sf9 cells. J Biol Chem 1992; 267:21733-7.

27. Morrison DF, O'Brien PJ, Pepperberg DR. Depalmitylation with hydroyxylamine alters the functional properties of rhodopsin. J Biol Chem 1991; 266:20118-23.

Typographical corrections