A Phe-rich region in short-wavelength sensitive opsins is responsible for their aggregation in the absence of 11-cis-retinal

Abstract

Human blue and mouse S-opsin are prone to aggregation in the absence of 11-cis-retinal, which underlie the rapid cone degeneration in human patients and animal models of Leber congenital amaurosis (LCA). By in silico analysis and domain swapping experiments, we show that a Phe-rich region in short-wavelength sensitive (SWS) opsins, but not in medium/long-wavelength sensitive opsins, is responsible for SWS opsin aggregation. Mutagenesis studies suggest that Phe residues in this region are critical in mediating protein aggregation. Fusing the Phe-rich region of SWS opsins to GFP causes the latter to aggregate. Our findings suggest that new therapeutics can be designed to disrupt the Phe-rich region in preventing cone degeneration due to S-opsin aggregation in LCA.

1. Introduction

Opsins are the light detector for all visual systems in the animal kingdom. They are members of a large super family of G-protein-coupled receptors (GPCRs) characterized by their seven transmembrane α-helix structure and their ability to covalently bind a vitamin A chromophore (e.g. 11-cis-retinal) [1]. Vertebrate opsins are integral membrane proteins located in the outer segments of photoreceptor cells (rods or cones) in the retina.

Both Retinoid isomerase (RPE65) and lecithin-retinol acyltransferase (LRAT) are required for 11-cis-retinal recycling in the RPE. Mutations in RPE65 or LRAT cause Leber congenital amaurosis (LCA), the most severe retinal dystrophy causing visual impairment in early childhood [2–5]. We have shown that mouse short-wavelength opsin (S-opsin) and human blue opsin are more prone to aggregation than the mouse M-opsin and human red/green opsins in the absence of 11-cis-retinal [6]. The accumulation of S-opsin triggers endoplasmic reticulum (ER) stress and rapid cone degeneration in the Lrat^−/− LCA model. This result explains the region-specific cone degeneration pattern in the retina of LCA mouse models since the ventral and central retina express higher levels of S-opsin than the dorsal retina in mice [7]. It also explains why blue-cone function is lost earlier than the red- and green-cone mediated function in human LCA patients [8–10].

It is unclear why short-wavelength sensitive (SWS) opsins (e.g. mouse S-opsin and human blue) are more susceptible to aggregation than medium- and long-wavelength sensitive (M/LWS) opsins (e.g. mouse M-opsin and human red/green) in the absence of 11-cis-retinal. Here we showed that a Phe-rich region in the SWS opsins, but absent from the M/LWS opsins, is responsible for the aggregation-prone property of SWS opsins.

2. Materials and Methods

2.1. Plasmids

Mouse M-opsin, mouse S-opsin, human red opsin, and human blue opsin in pRK5 were provided by Dr. Jeremy Nathans (Johns Hopkins University School of Medicine). A 1D4 tag (ETSQVAPA) was added to the C-terminal of various opsins for easy detection with the 1D4 monoclonal antibody. Plasmids encoding various chimera opsins (S^M86–141, M^S70–125, HB^HR91–146, and HR^HB72–127) were generated by PCR-based method. To mutate the 8 Phe residues of S-opsin, we synthesized the nucleotides encoding the Phe-rich region with mutated nucleotides underlined: 5’ ATT CTG GTC AAT GTA TCC CTC GGG GGC CTA CTC TTG TGC ATC ATT TCT GTC ACT ACA GTC GTT ATC GCC AGC TGT CAC GGA TAC TTC CTC CTG GGT CGC CAT GTT TGT GCT CTG GAG GCC TAC TTG GGC TCT GTA GCA GGT CTA GTG ACA GGA TGG TCA TTG GCT ATC 3’. The corresponding amino acid sequence is, 70 ILVNVSLGGLLLCIISVTTVVIASCHGYFLLGRHVCALEAYLGSVAGLVTGWSLAI 125. We then replaced the Phe-rich region of S-opsin with this synthesized region by overlap extension PCR. To construct plasmids encoding four fusion proteins (S^70–125-GFP, HB^72–127-GFP, M^86–141-GFP, and HR^91–146-GFP), we amplified the corresponding regions by PCR with the addition of the Kozak sequence and an ATG initiation codon on the 5’ end. The fragments were then cloned into pEGFP-N1vector (Clonetech) in frame with GFP by Nhe I and Bgl II.

2.2. Cell Culture

COS-7 cells were grown in DMEM supplemented with 10% FBS, 100 U/mL penicillin and 0.1 mg/mL streptomycin. Cells were seeded on cover slips in a 24-well plate. 24 hours after seeding, cells were transfected with 1 µg of various plasmids and 2 µL Turbofect reagent (Fermentas) following the manufacture’s instruction. Transfected cells were incubated for 48–55 hours before being fixed in 4% paraformaldehyde in PBS for 30 min at 37 °C. When we used the same amount of plasmid for different opsins, SWS opsin always had significant higher tendency to aggregate than medium/long-wavelength sensitive (M/LWS) opsins. We did not use 293 cells since these cells have functional retinoid processing machinery [11, 12].

2.3. Immunocytochemistry

Cells were permeated with 0.6% Triton X-100 in PBS for 30 min at room temperature after fixation. Except for cells transfected with plasmids expressing GFP fusion proteins, subcellular localization of opsins were detected with the 1D4 monoclonal antibody (1:100, hybridoma supernatant, from R.S. Molday) in Figs 2 & 3, and with the anti-mouse-S-opsin (MBO) (1:1000, from J. Chen) [13] in Fig. 4. The wells were blocked for 1 h in 5% goat serum in PBS. Cells were incubated with the primary antibody overnight at 4 °C followed by three washes in PBS and incubation of Alexa 488- or Cy3-conjugated secondary antibodies (1:500) for 1 h. After three washes in PBS, images were taken with Olympus Fluoview FV1000 confocal microscope.

Fig. 2.

Expression of chimera S^M86–141 and M^S70–125 opsins in COS-7 cells. (A) COS-7 cells were transfected with plasmids encoding various opsins, which were detected by immunocytochemistry (red). White arrows indicate intracellular inclusions formed from aggregated opsins. Scale bar, 10 µm. (B) Quantification of cells with aggregated opsins. The cells with aggregated opsins were counted per 100 transfected cells. N=6 for S, S^M86–141, M; N=5 for M^S70–125. ***, p<0.001, NS, not significant.

Fig. 3.

Expression of chimera HB^HR91–146 and HR^HB72–127 opsins in COS-7 cells. (A) Chimera opsins were detected by immunocytochemistry (red). White arrows indicate intracellular inclusions formed from aggregated opsins. Scale bar, 10 µm. (B) Quantification of cells with aggregated opsins. The cells with aggregated opsins were counted per 100 transfected cells. N=6 for HB and HR^HB72–127; N=8 for HB^HR91–146 and HR. ***, p<0.001, NS, not significant.

Fig. 4.

Expression of WT S-opsin (S^WT) and mutant S-opsin (S^mut) in COS-7 cells. (A) The 8 Phe residues in the Phe-rich regions of WT S-opsin, which were mutated in S^mut, are indicated. (B) WT and mutant S-opsin were detected by immunocytochemistry (green). White arrows indicate intracellular inclusions formed from aggregated opsins. Scale bar, 10 µm. (C) Quantification of cells with aggregated opsins. The cells with aggregated opsins were counted per 100 transfected cells. N = 6. ***, p<0.001, NS, not significant.

2.4. Statistics

Data were presented as mean ± SEM, and the differences were analyzed with unpaired two-sample Student’s t-test. P values < 0.05 were considered statistically significant.

3. Results

3.1. Identification of a Phe-rich region in the human blue and mouse S-opsin that is responsible for their aggregation-prone property

To identify regions that might be responsible for the aggregation-prone property of S-opsin, we analyzed the aggregation propensity profile with Zyggregator, a computer algorithm that predicts aggregation-prone regions of a protein based on the hydrophobic pattern, secondary structure propensity, and charges of amino acid sequences [14]. We identified a Phe-rich region at helix II (H-II)-extracellular loop I (E-I)-helix III (H-III) containing 9 Phe in human blue and mouse S-opsin, which has a high probability to aggregate (Fig. 1). In contrast, the homologous regions in human red/green and mouse M opsins contain only one Phe and are predicted to be much less likely to aggregate.

Fig. 1.

The aggregation-prone regions of S-opsin and human blue identified by Zyggregator are enriched with Phe. Regions in M-opsin and human red/green/blue opsins that are homologous to 70–125 of S-opsin are aligned. This region is located in helix II (H-II), extracellular loop I (E-I), and helix III (H-III). Phe residues are in bold. The secondary structures of cone opsins are modeled from the rhodopsin crystal structure [20]. “*”, identical; “:”, conserved substitutions; “.”, semi-conserved substitutions.

To determine the role of the Phe-rich region in mediating opsin aggregation, we generated plasmids encoding four chimera opsins, S^M86–141 (S opsin containing M-opsin region 86–141) and M^S70–125 (M opsin containing S-opsin region 70–125), HB^HR91–146 (human blue containing human red opsin region 91–146), and HR^HB72–127 (human red containing human blue opsin region 72– 127). We transfected COS-7 cells with plasmids encoding the four chimera opsins and the respective WT opsin controls. Opsin aggregation was determined by antibody labeling and confocal microscopy. S^WT and M^S70–125 opsins formed prominent aggregations manifesting as red dots of varying sizes (Fig. 2A, white arrows). This was not observed for M^WT and S^M86–141 opsins, which showed a more uniform distribution. The intracellular accumulation of both mouse S^WT and M^S70–125 was noticeably higher than that of M^WT and S^M86–141 opsins, even though the same amount of plasmid was used for transfection. M^WT and S^M86–141 opsins were most likely degraded as occurred in M^WT in Lrat^−/− photoreceptors [6]. Similarly, HB^WT and HR^HB72–127 had significantly higher tendency to aggregate than HR^WT and HB^HR91–146 (Fig. 3). Thus, domain swapping also swaps the aggregation property between SWS opsins and M/LWS opsins (Figs. 2 & 3), which is consistent with our in silico analysis (Fig. 1). This result suggests that the Phe-rich region in human blue and mouse S-opsin plays a major role in mediating their aggregation.

3.2. Mutation of Phe residues in the Phe-rich region of S-opsin prevents S-opsin aggregation in COS-7 cells

To investigate the role of the Phe residues of the S70–125 region in the different aggregation property between S-opsin and M-opsin, we mutated 8 Phe residues (Phe98 is identical to Phe114 in M-opsin) of S-opsin to the corresponding residues of M-opsin except that Phe81 was changed to Leu, which is a conservative change to maintain the hydrophobicity and size (Fig. 4A). In contrast to the prominent perinuclear aggregation in COS-7 cells transfected with WT S-opsin (Fig. 4B, left panel, white arrows), mutant S-opsin (S^mut) has a significantly reduced tendency to aggregate (Fig. 4B & C), suggesting Phe residues are critical in mediating S-opsin aggregation.

3.3. The Phe-rich region of human blue and mouse S-opsin mediates protein aggregation when fused to GFP

To test whether the Phe-rich region in mouse S-opsin and human blue is sufficient to mediate protein aggregation when added to a non-opsin “benign” protein (i.e. GFP), we constructed plasmids to express fusion proteins between the Phe-rich regions of mouse S-opsin or human blue with GFP in COS-7 cells (Fig. 5A). We also generated control plasmids to express fusion proteins between the homologous regions of mouse M-opsin and human red with GFP. In contrast to the diffused cytoplasmic distribution of GFP signal, S^70–125-GFP and HB^72–127-GFP showed prominent perinuclear aggregation (Fig. 5B, white arrows) whereas the distribution pattern of M^86–141-GFP and HR^91–146-GFP is similar to that of GFP (Fig. 5B). In combination with our domain swapping and mutagenesis data (Figs. 2–4), this result suggests that the Phe-rich region in human blue and mouse S-opsin is both necessary and sufficient to mediate protein aggregation.

Fig. 5.

Expression of S^70–125-GFP, M^86–141-GFP, HB^72–127-GFP, and HR^91–146-GFP fusion proteins in COS-7 cells. (A) Schematic drawing on the construction of fusion proteins between GFP and the Phe-rich regions of mouse S-opsin/human blue or the homologous regions in human red and green opsins. (B) Expression of various GFP-fusion proteins (green). White arrows indicate intracellular inclusions formed from aggregated fusion proteins. Scale bar, 10 µm.

4. Discussion

In this study, our domain swapping (Figs. 2 & 3) and fusion protein (Fig. 5) experiments suggest that the Phe-rich region in S-opsin is responsible for its aggregation in the absence of 11-cis-retinal (e.g. in Lrat^−/− retina). Our mutagenesis study suggests that a high concentration of Phe residues in this region is critical in mediating the self-aggregation process of S-opsin. To the best of our knowledge, this is the first demonstration that such a pathophysiological mechanism exists in the opsin family in causing blinding diseases.

SWS1 opsins exhibit the shortest wavelength sensitivity among vertebrate opsins with peaks in either the ultraviolet (around 365 nm) or violet (390–435 nm) in different species [15]. When we compared the homologous regions of S-opsin (70–125) in the SWS1 family from 64 species, we found that the “Phe-rich” feature was well conserved in mammals, birds, reptiles, and amphibians (Table 1, Fig. S1). Many fishes tend to have less Phe (5–6), which seem to be compensated by an increased number of Tyr residues, making the total number of aromatic residues relatively constant (Table 1). In sharp contrast, the number of Phe is much lower (1–2) in the M/LWS opsin family from mammals, birds, and amphibians (Table 2, Fig. S2). Some species of fish and reptiles have slightly more Phe (3–5). Out of the three aromatic residues, Trp shows the least variation between SWS1 and M/LWS opsins, containing only one highly conserved Trp (corresponding to W121 in S-opsin) except that cavefish MWS and the African clawed frog SWS1 opsin have two. The number of Tyr ranges between one and three in both SWS1 and M/LWS. Phe shows the biggest variation between the two opsin families, which may underlie their very different aggregation property in the absence of 11-cis-retinal. One implication of our study is that the aggregation-prone property is a common feature of the SWS1 opsins. This is generally not a problem for SWS1 opsins under normal condition when 11-cis-retinal is continuously regenerated in the eye. A recent study suggests that 11-cis-retinal binds to cone opsins during biosynthesis in the ER to help folding [16]. Thus, 11-cis-retinal is essential for preserving a transport-competent conformation of cone opsins and preventing aggregate formation, which is achieved by shielding hydrophobic surfaces (e.g. 70–125 in S-opsin) from forming inappropriate intermolecular contacts during protein synthesis and transport (i.e. as a chemical chaperone). Indeed, seven out of the nine Phe residues of S-opsin are located in the transmembrane regions (H-II & H-III) (Fig. 1), which are protected by the lipid bilayer in the outer segment of cone photoreceptors after correct targeting.

Table 1.

Comparison of the number of aromatic residues (F, W, Y) in the homologous region of the phenylalanine-rich region of S-opsin (70–125) in the SWS1 opsin family.

Class	Speciesa	F	W	Y	Total
Bony fish	goldfish	7	1	3	11
	zebrafish	5	1	3	9
Amphibian	salamander	9	1	1	11
	clawed frog	9	1	1	11
Reptile	day gecko	9	1	1	11
	lizard	11	1	1	13
Bird	pigeon	9	1	1	11
	zebra finch	8	1	1	10
	human	9	1	1	11
	macaque	10	1	1	12
Mammal	cow	8	1	3	12
	pig	8	1	2	11
	mouse	9	1	1	11

^aOnly representative species are shown here due to space limitation. A more exhaustive comparison of 64 species is shown in Figure S1 in Supplementary Material.

Table 2.

Comparison of the number of aromatic residues (F, W, Y) in the homologous region of mouse M-opsin (86–141), which corresponds to the phenylalanine-rich region of S-opsin (70–125), in the M/LWS opsin family.

Class	Speciesa	F	W	Y	Total
Bony fish	Goldfish LWS	5	1	1	7
	Zebrafish LWS	4	1	2	7
	Cavefish MWS	3	2	1	6
Amphibian	Salamander LWS	2	1	2	5
	X. Tropicalis LWS	2	1	2	5
Reptile	Gecko MWS	4	1	2	7
	Snake LWS	3	1	2	6
	Lizard LWS	1	1	2	4
Bird	Pigeon LWS	1	1	2	4
	Chicken LWS	1	1	2	4
	Canary LWS	2	1	2	5
	Human LWS	1	1	2	4
	Macaque LWS	2	1	1	4
	Whale LWS	2	1	2	5
	Boar MWS	1	1	3	5
Mammal	Goat LWS	1	1	3	5
	Cow LWS	1	1	3	5
	Mouse MWS	1	1	3	5

^aOnly representative species are shown here due to space limitation. A more exhaustive comparison of 37 species is shown in Figure S2 in Supplementary Material.

The high percentage of Phe residues in the Phe-rich region of SWS1 opsins (16–20%) is surprising because the frequency of occurrence of aromatic residues in proteins is generally not high (3.9% for Phe) [17, 18]. Aromatic residues are highly conserved in proteins, suggesting important roles in structural biology [19]. It is unlikely that so many Phe residues are needed to maintain the secondary or tertiary structures of SWS1 opsins because SWS1 and M/LWS opsins are expected to share similar structures [20]. More likely, the Phe residues are involved in spectral tuning or the specialized function of SWS1 opsins in detecting ultraviolet or violet light. For example, Phe81 in S-opsin is important for UV spectral tuning [21–23]. More studies are necessary to clarify the function of the large number of Phe residues in this region of the SWS1 opsins.

Aromatic amino acids strongly promote amyloid formation in a number of conformational diseases [24–26]. The high amyloidogenicity is mainly attributed to the high hydrophobicity, β-sheet propensity, and planar geometry of aromatic residues although π-π interactions may also be involved in some amyloid peptides [24, 25, 27, 28]. The fact that S^mut protein is much less likely to aggregate than S-opsin (Fig. 4) suggests that aromatic amino acids (i.e. Phe) are critical in mediating S-opsin self-aggregation. We propose that the large number of Phe residues in the Phe-rich region in SWS1 opsins mediate opsin aggregation via a similar mechanism as in other conformational diseases. Our study has important implications for the therapeutic strategy for LCA. Recently, gene therapy delivering AAV2-RPE65 by subretinal injection offers great promise with successful Phase I and II clinical trials [29–34]. However, gene therapy has a limited time window, i.e. before significant retinal degeneration occurs [32, 33]. This is especially true in rescuing cones. Jacobson, S.G. et al concludes that there is no benefit and some risk in treating the cone-rich fovea [34]. Moreover, a recent study showed that gene therapy improves vision for at least 3 years in LCA patients, but photoreceptor degeneration progresses unabated [35]. This study shows the need for combinatorial therapy to improve vision and to slow retinal degeneration. Potentially, novel compounds that can disrupt the interaction between the Phe-rich region of S-opsin may be used to prevent or slow cone degeneration.

Highlights.

• □A Phe-rich region in short-wavelength sensitive (SWS) opsins (i.e. mouse S-opsin and human blue) is identified by in silico analysis and confirmed by domain swapping experiments to be responsible for their aggregation prone property.
• □Mutagenesis studies suggest Phe residues play a key role in mediating SWS opsin aggregation.
• □Fusing the Phe-rich region of SWS opsins to GFP causes the latter to aggregate.
• □The Phe-rich region is conserved in all the SWS1 opsin family in vertebrates (64 species compared).