Mutations that perturb the function of photoreceptor CNG (cyclic nucleotide-gated) channels are associated with several human retinal disorders, but the molecular and cellular mechanisms leading to photoreceptor dysfunction and degeneration remain unclear. Many loss-of-function mutations result in intracellular accumulation of CNG channel subunits. Accumulation of proteins in the ER (endoplasmic reticulum) is known to cause ER stress and trigger the UPR (unfolded protein response), an evolutionarily conserved cellular programme that results in either adaptation via increased protein processing capacity or apoptotic cell death. We hypothesize that defective trafficking of cone photoreceptor CNG channels can induce UPR-mediated cell death. To test this idea, CNGA3 subunits bearing the R563H and Q655X mutations were expressed in photoreceptor-derived 661W cells with CNGB3 subunits. Compared with wild-type, R563H and Q655X subunits displayed altered degradation rates and/or were retained in the ER. ER retention was associated with increased expression of UPR-related markers of ER stress and with decreased cell viability. Chemical and pharmacological chaperones {TUDCA (tauroursodeoxycholate sodium salt), 4-PBA (sodium 4-phenylbutyrate) and the cGMP analogue CPT-cGMP [8-(4-chlorophenylthio)-cGMP]} differentially reduced degradation and/or promoted plasma-membrane localization of defective subunits. Improved subunit maturation was concordant with reduced expression of ER-stress markers and improved viability of cells expressing localization-defective channels. These results indicate that ER stress can arise from expression of localization-defective CNG channels, and may represent a contributing factor for photoreceptor degeneration.

INTRODUCTION

CNG (cyclic nucleotide-gated) channels are critical to the phototransduction cascade that underlies vertebrate vision. These channels open and close in response to light-induced changes in the intracellular cGMP concentration; the resulting changes in membrane potential modulate activity at the photoreceptor synaptic terminal (reviewed in [1]). In mammals, six genes encode two subunit types (A and B), which form heterotetrameric channels. Mutations in cone photoreceptor CNG channel genes are associated with complete and incomplete ACHM (achromatopsia), PCD (progressive cone dystrophy) and MD (macular degeneration) [25]. Forms of ACHM, PCD and MD linked to mutations in CNGA3 and CNGB3 are characterized by intact rod function and limited or absent cone function. ACHM, once considered a stable cone dystrophy, has been shown previously to exhibit progressive cone photoreceptor loss in some patients [2,3], whereas PCD and MD are distinguished by cone degeneration [4,5]. Genetic and clinical heterogeneity of these disorders has complicated phenotype–genotype correlations [4,6], and the underlying mechanisms causing photoreceptor dysfunction and death in response to CNG channel mutations are not well understood. Functional characterization of CNG channels bearing disease-associated mutations has revealed both gain- and loss-of-function phenotypes [712]. Reduced or absent PM (plasma membrane) localization commonly contributes to the functional deficit of loss-of-function mutations [7,1012]. As a prerequisite to assuming their appropriate PM location, CNG channels must undergo a number of protein maturation steps, including folding, assembly and trafficking out of the ER (endoplasmic reticulum). Disposal of proteins that cannot successfully mature imposes a metabolic burden on the cell that may contribute to the pathophysiology associated with mutations that give rise to localization-defective channels. For such CNG channel mutations, whether the loss of channel function is the sole pathogenic culprit, or if the metabolic stress of processing defective proteins also plays a role in disease progression remains unexplored.

Accumulation of proteins in the ER activates the UPR (unfolded protein response), a co-ordinated signalling programme that protects the cell from ER stress (reviewed in [13]). Three resident ER luminal transmembrane proteins serve as proximal sensors of ER stress and initiate eponymous branches of the UPR: PERK [PKR (double-stranded-RNA-dependent protein kinase)-like ER kinase], IRE1 (inositol-requiring enzyme 1) and ATF6 (activating transcription factor 6). Activated PERK phosphorylates eIF2a (eukaryotic translation initiating factor 2a), rendering it unable to initiate global mRNA translation and immediately reducing the processing load of the ER. Activation of IRE1 and ATF6 up-regulates expression of cytoprotective proteins involved in the processing capacity of the ER, including chaperone proteins [e.g. BiP (immunoglobulin heavy-chain-binding protein) and the transcription factor XBP1 (X-box-binding protein 1)]. Paradoxically, ER stress also induces production of pro-apoptotic signals {e.g. CHOP [C/EBP (CCAAT/enhancer-binding protein)-homologous protein]}. The speed, magnitude and maintenance of specific signals determine the balance between adaptive and apoptotic pathways, neither of which is exclusive to any branch of the UPR [14]. If the cell cannot prevent accumulation of misfolded proteins, the prolonged perturbation from ER homoeostasis can result in apoptotic cell death (reviewed in [15]).

ER stress has been linked to several channelopathies associated with channel mutations that cause trafficking defects, including the I593R mutation in the HERG (human ether-a-go-go-related gene) associated with LQT2 (long QT syndrome type 2) [16] and the ΔF508 mutation in the CFTR [CF (cystic fibrosis) transmembrane conductance regulator], the most common cause of CF [17]. Chemical chaperones, agents that intervene in the UPR to reduce ER stress, have been shown to improve trafficking of CFTR ΔF508 [18]. Furthermore, the function of a number of HERG trafficking-defective channels (e.g. N470D) can be rescued by channel blockers acting as pharmacological chaperones [19]. Such agents bind specific structural elements within the substrate protein and reduce ER stress by promoting maturation, degradation and/or trafficking of defective proteins [2022]. The relative importance of channel activity restoration compared with ER-stress reduction remains undetermined for these channelopathies.

The involvement of the UPR in the pathophysiology of retinal degenerative disorders is also becoming evident [2325]. Transgenic expression of autosomal-dominant RP (retinitis pigmentosa)-associated P23H rhodopsin, known to accumulate in the ER and form pericentriolar inclusion bodies, induced ER stress in rat retinas [26]. Addition of the pharmacological chaperone and chromophore 11-cis-retinal promoted maturation and PM localization of P23H rhodopsin when heterologously expressed in HEK (human embryonic kidney)-293 cells [21,22]. In mouse cone photoreceptor-derived 661W cells, co-expression of putative chaperone proteins reduced ER accumulation of P23H rhodopsin [27]. Recently, visual function was restored in P23H transgenic rats using viral gene delivery of the endogenous chaperone BiP [25]. Despite gathering evidence implicating ER stress in retinal degeneration, this issue remains unexamined in the context of mutant CNG channels that fail to localize to the PM.

In the present study, we test the hypothesis that trafficking-defective CNG channel subunits cause ER stress and UPR-associated cell death. We observed changes in markers of ER stress and cell viability in photoreceptor-derived cells after heterologous expression of localization-defective CNGA3 R563H and Q655X subunits, which have been previously linked to PCD and complete ACHM [6]. Chemical and pharmacological chaperones were found to relieve ER stress and cytotoxicity, and to improve stability, ER exit and/or PM localization of CNG channel subunits. The results of the present study suggest a connection between expression of localization-defective channel subunits and UPR-mediated cytotoxicity.

EXPERIMENTAL

Chemicals

The chemicals used were from the following sources: DTT (dithiothreitol) (catalogue number 27565-41-9, Roche Diagnostics), Tn (tunicamycin) (catalogue number 11089-65-9, Calbiochem/EMD Biosciences), Tg (thapsigargin) (catalogue number 67526-95-8, Sigma–Aldrich), TUDCA (tauroursodeoxycholate sodium salt) (catalogue number 14605-22-2, Calbiochem), 4-PBA (sodium 4-phenylbutyrate) (catalogue number 1716-12-7, Calbiochem), CPT-cGMP [8-(4-chlorophenylthio)-cGMP] (catalogue number 51239-26-0, Sigma–Aldrich), cycloheximide (catalogue number 66-81-9, Calbiochem), trypsin (catalogue number 9002-07-7 Sigma–Aldrich) and lima bean trypsin inhibitor (catalogue number 9035-81-8, Sigma–Aldrich).

Molecular biology, cell culture and transfection

For expression in mammalian cells, the coding sequences for wild-type human CNGA3 (GenBank® accession number Q16281) and CNGB3 (GenBank® accession number Q9NQW8) subunits [9,28] were fused with an N-terminal EGFP [enhanced GFP (green fluorescent protein)] or triple-FLAG tag and subcloned into the pcDNA3.1/Zeo(+) plasmid (Invitrogen) using unique restriction sites. Point mutations in A3 (R563H) and B3 (F525N) were made using the QuikChange® II Site-Directed Mutagenesis kit (Stratagene); the Q655X deletion was made by PCR cassette mutagenesis. The murine cone photoreceptor-derived 661W cell line was a gift from Professor M.R. Al-Ubaidi (University of Oklahoma Health Sciences Center, Oklahoma City, OK, U.S.A.). Cells were maintained in DMEM (Dulbecco's modified Eagle's medium; Gibco) supplemented with 10% FBS (fetal bovine serum; Gemini Bioproducts), penicillin (100 units/ml) and streptomycin (100 mg/ml; Gibco) at 37°C in a humidified 5% CO2 incubator. For pharmacological ER-stress induction, cells were seeded in six-well (10 cm2/well) plates at a density of approximately 2×105 cells/well in medium supplemented with 2 mM DTT, 10 μg/ml Tn or 0.5 μM Tg. At intervals after the addition of chaperone agents, cells from duplicate wells were harvested simultaneously for RNA or protein for each treatment. Untreated control cells were harvested at zero time. For CNG channel experiments, cells were co-transfected with pcDNA3.1/Zeo(+) plasmids encoding CNGA3 and CNGB3 subunits at a ratio of 1:1 (to form heteromeric channels) using Lipofectamine™ 2000 and OptiMEM (Invitrogen) according to the manufacturer's protocol for cells in suspension. For RNA and protein experiments, cells were plated in six-well (10 cm2/well) plates at a density of approximately 1.6×105 cells/well. For LDH (lactate dehydrogenase) assays, cells were plated in 96-well (0.3 cm2/well) plates at a density of approximately 8×103 cells/well. Transfection efficiency was assessed by GFP expression from the reporter plasmid pQBI25-fC2 (Qbiogene) in a transfection control sample; transfection efficiency was typically ~75%. The pcDNA3.1/Zeo(+) plasmid was used as a control to determine transfection toxicity. The pDsRed2-ER plasmid (Clontech) was used to identify the ER in confocal microscopy experiments. For chaperone treatments, medium was exchanged at 3 h post-transfection for DMEM supplemented with 10% FBS with or without chaperone agents (1 mM TUDCA, 1 mM 4-PBA or 1 μM CPT-cGMP. For CNG channel subunit stability experiments, cycloheximide was added to a concentration of 75 μg/ml at 40 h post-transfection.

RT–qPCR (reverse transcription–quantitative PCR)

At 24 h post-transfection, RNA was extracted and purified using TRIzol® (Invitrogen) according to the manufacturer's protocol. Oligo(dT)-primed RT was performed using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen). Platinum Quantitative PCR SuperMix-UDG (Invitrogen) was used to perform qPCR. Briefly, 25 μl reaction volumes (1 ng of cDNA and 10 μM primers) were run in triplicate in 96-well plates as follows: 2 min at 50°C to degrade contaminants, 2 min at 95°C to deactivate UDG and activate Platinum Taq, followed by 40 cycles of 15 s at 95°C (melting), 15 s at 58°C (annealing) and 15 s at 72°C (elongation). An iCycler version 2.039 thermocycler (Bio-Rad Laboratories) was used, and threshold cycles (Ct) were determined using iCycler IQ Optical System Software version 3.1 set for baseline subtracted curve fit. Temperature gradient reactions were used to confirm efficient priming at 58°C for all primer sets. The specificity of priming was confirmed by melt curve and electrophoretic analysis of products. Primer efficiency was calculated after determining Ct for the template concentrations over at least three orders of magnitude. ΔCt between treated and untreated cells was determined for all genes (except CNG channels), and the Pfaffl Method [29] was used to calculate fold changes using HPRT (hypoxanthine–guanine phosphoribosyltransferase) to standardize the amount of RNA [30] in samples. 661W cells do not express detectible levels of CNGA3 mRNA [31] and therefore fold changes cannot be anchored, thus ΔCt between CNGA3 and HPRT were compared directly between samples. The percentage of spliced XBP1 was determined using the ratio of fold-changes for spliced XBP1/total XBP1.

Primers

The primers used in the present study were as follows (m indicates mouse, h indicates human): mHPRT (sense) 5′-AGTCCCAGCGTCGTGATTAG-3′; mHPRT (antisense) 5′-CCAGCAGGTCAGCAAAGAAC-3′; spliced mXBP1 (sense) 5′-TGAGTCCGCAGCAGGTG-3′; spliced mXBP1 (antisense) 5′-GAGGCAACAGTGTCAGAGTCC-3′; mXBP1 (sense) 5′-AGAACACGCTTGGGAATGG-3′; mXBP1 (antisense) 5′-CCACCAGCCTTACTCCACTC-3′; mBiP (sense) 5′-TGTTTGTCCCCTTACACTTGG-3′; mBiP (antisense) 5′-GGCGGTTTTGGTCATTGG-3′; mCHOP (sense) 5′-GTCCCTGCCTTTCACCTTG-3′; mCHOP (antisense) 5′-TCGTTCTCCTGCTCCTTCTC-3′; hCNGA3 (sense) 5′-ATCACCTCGTGTGTTCTTTGG-3′; hCNGA3 (antisense) 5′-GGCTGTCTTTGAATCTCTTTG-3′; hCNGB3 (sense) 5′-AGCAACCCCTCCAAGAAAAG-3′; hCNGB3 (antisense) 5′-GGGGTTCTTCCTCCACTGC-3′.

Immunoblotting

Cells were pelleted by centrifugation at 300 g for 3 min, gently washed with PBS, and resuspended in 20 μl of ice-cold lysis buffer containing 20 mM Tris/HCl (pH 7.4), 150 mM sodium chloride, 1 mM EGTA, 1 mM sodium fluoride, 1 mM sodium orthovanadate, 1% Nonidet P40 and 0.25% sodium deoxycholate, plus one protease inhibitor cocktail tablet (Complete™ Mini EDTA-free; Roche Applied Science) per 50 ml of lysis buffer. To each sample, 2 units of Turbo DNAse (Ambion) were added. Samples were placed on ice and allowed to lyse for 20 min. Cellular debris was pelleted by centrifugation at 5000 g for 5 min, followed by collection of the supernatants. NuPAGE® LDS Sample Buffer and Reducing Agent (Invitrogen) were immediately added, and samples were heated to 70°C for 10 min prior to gel loading. Proteins were separated by SDS/PAGE using 4–12% Bis-Tris NuPAGE® gels in Mes/SDS Running Buffer plus Antioxidant (Invitrogen), then transferred on to nitrocellulose membranes using NuPage® Transfer Buffer (Invitrogen). Blots were incubated for 1 h in 50% Odyssey Blocking Buffer (LI-COR Biosciences) in TBS [Tris-buffered saline; 50 mM Tris/HCl (pH 7.4) and 150 mM NaCl], then overnight in primary antibody (or 1 h for anti-actin and anti-CHOP antibodies) in 50% Odyssey Blocking Buffer in TBS plus 0.1% Tween 20 on ice. Blots were washed extensively in TBS plus 0.1% Tween 20. Blots were incubated for 2 h in secondary antibodies in 50% Odyssey Blocking Buffer in TBS with 0.1% Tween 20 and 0.01% SDS on ice, then washed extensively. Proteins were detected using the Odyssey Infrared Imaging System and Software version 2.0 (LI-COR Biosciences) at medium resolution and 5.0 intensity. Protein bands were analysed and quantified using LI-COR Software; values were normalized to actin for each lane.

Cell-surface biotinylation

At 48 h post-transfection, cells were gently rinsed three times with chilled (4°C) DPBS (Dulbecco's PBS) (pH 8.0) to remove media, and then treated with 0.01 mg/ml SulfoLink NHS-LC-biotin [succinimidyl-6-(biotinamido)hexanoate; Pierce] in chilled DPBS (pH 8.0) for 30 min with gentle rocking at 4°C. Reactions were terminated by washing cells three times for 5 min with ice-cold 100 mM glycine in PBS (pH 8.0). Cells were gently scraped from the plate, transferred to tubes and harvested for immunoblotting as described above.

Immunohistochemistry and confocal microscopy

Transfected cells were grown on poly-L-lysine-coated glass coverslips. At 48 h post-transfection, cells were rinsed with PBS, and then fixed with 4% paraformaldehyde in PBS for 6 min. After three 10 min washes in PBS, cells were incubated with Alexa Fluor®-conjugated WGA (wheat germ agglutinin) in PBS for 10 min, then rinsed twice for 10 min with PBS. Cells were then permeabilized for 30 min in 0.5% Triton X-100, 1% BSA, and 5% NGS (normal goat serum). Cells were rinsed three times with PBS, and then incubated for 1 h in 0.1% Triton X-100, 1% BSA, 5% NGS, and appropriate dilutions of primary antibodies. Cells were washed three times in PBS, and then incubated in 0.1% Triton X-100, 1% BSA, 5% NGS and appropriate secondary antibodies at 1:2000. After three 10 min washes in PBS, coverslips were mounted on to slides with Prolong Gold AntiFade Reagent (Molecular Probes). Images were collected as 0.4–0.6 μm Z-sections using an Olympus FluoView 1000 laser-scanning microscope in XYZ mode with a 60×/1.4 NA (numerical aperture) oil-immersion lens. Line scan intensity and co-localization calculations were performed using FluoView 10-ASW version 1.6 software.

Membrane isolation and protease digestion

Parallel plates of similarly confluent cultures were used to isolate crude membrane fractions. All steps were performed at 4°C. Briefly, adherent cells were rinsed with PBS and scraped off into a 2 ml solution of 200 mM sodium chloride, 33 mM sodium fluoride, 10 mM EDTA and 50 mM Hepes (pH 7.4 with NaOH), plus a protease inhibitor cocktail (100 μM PMSF, 1 μg/ml pepstatin A, 1 μg/ml of leupeptin and 4 μl/ml of aprotinin). The cells were homogenized and centrifuged at 500 g for 10 min. The membrane fractions were pelleted from the low-speed supernatants by centrifugation at 60000 rev./min in a TLA100.1 rotor (Beckman) for 1 h at 4°C. Membrane pellets were resuspended in 50 μl of TBS. For each protease digest time course, 25 ng of trypsin was added to 70 μl of a 10-fold membrane dilution. Time point samples of 10 μl were transferred to microfuge tubes containing 2 μg of trypsin inhibitor. At end of the time course, samples were analysed by immunoblotting.

Antibodies

Antibody sources, dilutions, and applications were as follows: anti-phospho-PERK (rabbit monoclonal, 1:500 dilution; Cell Signaling Technology), anti-GRP78 BiP (rabbit polyclonal, 1:5000 dilution; Abcam), anti-GADD 153 (CHOP) (mouse monoclonal, 1:1000 dilution; Santa Cruz Biotechnology), anti-actin (mouse monoclonal, 1:5000 dilution; Chemicon/Millipore), anti-FLAG M2 (mouse monoclonal, 1:50000 dilution for immunoblot analysis or 1:5000 dilution for confocal microscopy; Sigma–Aldrich), anti-rabbit (IR680) (donkey, 1:5000 dilution; LI-COR Biosciences), anti-mouse (IR800) (donkey, 1:10000 dilution; LI-COR Biosciences), Alexa Fluor® 495-conjugated WGA (at 5 μg/ml; Invitrogen), anti-DsRed (rabbit, 1:500 dilution; Clontech), anti-mouse Alexa Fluor® 488-conjugated (goat; Invitrogen), anti-rabbit Alexa Fluor® 594-conjugated (goat; Invitrogen) and anti-rabbit Alexa Fluor® 647-conjugated (donkey; Invitrogen).

Cell viability assay

An LDH cytotoxicity assay kit (Roche Applied Science) was used according to the the manufacturer's instructions. Briefly, cultured cells were transfected as described above and plated in 96-well tissue culture plates. At 3 h post-transfection, medium was exchanged for DMEM supplemented with 1% FBS with or without chaperone agents, with five to ten technical replicates per condition. After 48 h of incubation, half of the culture medium in each well was transferred to the identical well in another 96-well plate. The cells in the original plate were lysed, releasing LDH from intact cells into the remaining half of the culture medium. LDH activity was measured in both plates as per the protocol using a coupled enzymatic reaction that uses LDH-reduced NADH to reduce the tetrazolium salt to formazan, which absorbs light at 490 nm. LDH from dead cells was determined by doubling the amount measured in the medium plate. LDH in intact cells was determined by subtracting the LDH measured in the medium plate from the LDH measured in the original (lysed cells) plate. The percentage cytotoxicity was calculated from the ratio of LDH in the medium/cells, and was normalized to the percentage cytotoxicity in cells transfected with empty vector. Chaperone rescue was assessed by comparing the percentage cell death of chaperone-treated and untreated cells expressing the same CNGA3 subunit.

Statistical analysis

Data were analysed using SigmaPlot and SigmaStat (Systat Software), and expressed as means±S.E.M. Statistical significance was determined using ANOVA or paired (by biological replicate) Student's t test; P<0.05 were considered significant.

RESULTS

Characterization of disease-associated R563H and Q655X CNGA3 mutations

Several mutations in the human CNGA3 channel subunit have been associated with retinal disease. The R563H mutation in CNGA3 lies within the CNBD (cyclic nucleotide-binding domain) (Figure 1A) and alters cell-surface localization [7], whereas the Q655X (Q655stop) mutation disrupts a region important for CNG channel assembly [32], but little is known about the mechanistic underpinnings of the disease pathology. To investigate these mechanisms, we transiently expressed FLAG epitope-tagged CNGA3 and FLAG- or EGFP-tagged CNGB3 channel subunits in cone photoreceptor-derived 661W cells. These cells express elements of the phototransduction cascade [31,33]. The absence of endogenous CNGA3 and minimal expression of CNGB3 channel subunits [31] make 661W cells a particularly useful model for the study of exogenous CNG channels in a cone photoreceptor-derived cellular environment. We determined transient expression levels of FLAG-tagged CNGA3 subunits and found that the mutations did not significantly alter steady-state levels relative to wild-type subunits (Figure 1B). To quantify cell-surface localization of CNGA3 subunits, we employed surface biotinylation using membrane-impermeable reagents (Figure 1B). Relative to wild-type, surface localization of R563H and Q655X was significantly reduced (P<0.05 and P=0.02 respectively) (Figure 1C). Using confocal fluorescence microscopy, we observed significantly reduced PM localization of R563H and Q655X relative to wild-type CNGA3 in 661W cells (P<0.006) (Figures 1D and 1E). Next, we co-expressed an ER marker protein (Discosoma sp. red fluorescent protein DsRed2 fused to the ‘KDEL’ ER-retention sequence) and evaluated co-localization with CNGA3 subunits using confocal microscopy (Figure 1F). Compared with wild-type subunits, R563H and Q655X showed significantly greater ER localization (P<0.001) (Figure 1G). These results demonstrate that CNGA3 R563H and Q655X mutations reduced the trafficking competency and increased ER retention of CNGA3 subunits in 661W cells.

Disease-associated mutations R563H and Q655X impair PM localization and increase ER retention of CNGA3 channel subunits

Figure 1
Disease-associated mutations R563H and Q655X impair PM localization and increase ER retention of CNGA3 channel subunits

(A) CNGA3 subunit topology and location of mutations. (BG) Expression and localization of human cone CNGA3 and CNGB3 subunits in 661W cells, determined using cell-surface biotinylation with avidin-retrieval (B and C) or confocal microscopy (DG). (B) Representative immunoblot showing total (1% of input, with actin for normalization) and surface-localized CNGA3 subunits. (C) Cell-surface localization of CNGA3 subunits, displayed as avidin pullouts from (B) relative to the wild-type, is impaired by R563H and Q655X mutations. Results are means±S.E.M. from six independent experiments; *P<0.05. (D) Confocal fluorescence microscopy images and graphs showing PM marked with WGA (red) and FLAG–CNGA3 (green); PM localization (yellow) of R563H and Q655X subunits is impaired. (E) Co-localization calculated for regions encompassing cross-sections in (D) using Pearson's product-moment correlation coefficient, showing decreased PM localization of R563H and Q655X subunits. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=6); *P<0.0001. (F) Confocal images showing CNGA3 subunits (green) and ER marked with DsRed2-ER (blue). (G) Co-localization calculated for regions defined in (F) using Pearson's product-moment correlation coefficient, showing increased ER localization of R563H and Q655X subunits. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=6); *P<0.05. wt, wild-type.

Figure 1
Disease-associated mutations R563H and Q655X impair PM localization and increase ER retention of CNGA3 channel subunits

(A) CNGA3 subunit topology and location of mutations. (BG) Expression and localization of human cone CNGA3 and CNGB3 subunits in 661W cells, determined using cell-surface biotinylation with avidin-retrieval (B and C) or confocal microscopy (DG). (B) Representative immunoblot showing total (1% of input, with actin for normalization) and surface-localized CNGA3 subunits. (C) Cell-surface localization of CNGA3 subunits, displayed as avidin pullouts from (B) relative to the wild-type, is impaired by R563H and Q655X mutations. Results are means±S.E.M. from six independent experiments; *P<0.05. (D) Confocal fluorescence microscopy images and graphs showing PM marked with WGA (red) and FLAG–CNGA3 (green); PM localization (yellow) of R563H and Q655X subunits is impaired. (E) Co-localization calculated for regions encompassing cross-sections in (D) using Pearson's product-moment correlation coefficient, showing decreased PM localization of R563H and Q655X subunits. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=6); *P<0.0001. (F) Confocal images showing CNGA3 subunits (green) and ER marked with DsRed2-ER (blue). (G) Co-localization calculated for regions defined in (F) using Pearson's product-moment correlation coefficient, showing increased ER localization of R563H and Q655X subunits. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=6); *P<0.05. wt, wild-type.

Proteins that fail to fold properly are retained in the ER, where they accumulate or proceed to ERAD (ER-associated degradation) (reviewed in [13]). Thus we investigated the effects of R563H and Q655X mutations on the stability of CNGA3 and co-expressed CNGB3 subunits. After treatment with cycloheximide to halt protein synthesis, the amount of CNGA3 remaining in 661W cells was quantified over the course of 24 h. Core-glycoslated and unglycosylated subunits (Figure 2A, top and bottom bands respectively) were observed and their intensities pooled. The degradation rate relative to wild-type CNGA3 was increased by the Q655X mutation, but unaffected by the R563H mutation (Figures 2A and 2C). Simultaneously, the degradation rate of CNGB3 subunits was increased when co-expressed with R563H or Q655X, compared with expression with wild-type CNGA3 (Figures 2B and 2D). These results suggest that mutations in CNGA3 can give rise to structural abnormalities subject to the quality-control functions of the ER.

Localization-defective channels display altered stability and protease sensitivity

Figure 2
Localization-defective channels display altered stability and protease sensitivity

(AD) Amount of CNGA3 and CNGB3 subunits remaining at the indicated time points after arrest of protein synthesis. (A, B) Representative immunoblots of CNGA3 or CNGB3 subunits respectively. (C) Time-dependent decrease of CNGA3 subunits expressed as a percentage of the initial amount, showing increased degradation rate for Q655X relative to wild-type. (D) The percentage of the initial amount of CNGB3, with the indicated CNGA3 subunits, showing that trafficking-defective CNGA3 subunits increased the degradation rate of co-expressed CNGB3 subunits. Results are means±S.E.M. for seven independent experiments. *P<0.05. (E and F) Protease-sensitivity assay to probe CNGA3 subunit folding. (E) Representative Western blot for proteolysis of purified channels. (F) The percentage of the initial amount of CNGA3 remaining after trypsin proteolysis at the indicated time points, showing altered trypsin sensitivity. Results are means±S.E.M. for three independent experiments; *P<0.05.

Figure 2
Localization-defective channels display altered stability and protease sensitivity

(AD) Amount of CNGA3 and CNGB3 subunits remaining at the indicated time points after arrest of protein synthesis. (A, B) Representative immunoblots of CNGA3 or CNGB3 subunits respectively. (C) Time-dependent decrease of CNGA3 subunits expressed as a percentage of the initial amount, showing increased degradation rate for Q655X relative to wild-type. (D) The percentage of the initial amount of CNGB3, with the indicated CNGA3 subunits, showing that trafficking-defective CNGA3 subunits increased the degradation rate of co-expressed CNGB3 subunits. Results are means±S.E.M. for seven independent experiments. *P<0.05. (E and F) Protease-sensitivity assay to probe CNGA3 subunit folding. (E) Representative Western blot for proteolysis of purified channels. (F) The percentage of the initial amount of CNGA3 remaining after trypsin proteolysis at the indicated time points, showing altered trypsin sensitivity. Results are means±S.E.M. for three independent experiments; *P<0.05.

To investigate the effects of these mutations on the folding of CNGA3 subunits, we isolated membranes from cells expressing wild-type, R563H or Q655X subunits and subjected them to trypsin proteolysis (Figure 2E). Changes in proteolytic rate are presumed to indicate changes in the accessibility of trypsin-labile sites, which is used as an indicator of changes in protein structure. Recapitulating the stability results, trypsin proteolysis of R563H was indistinguishable from wild-type, whereas Q655X showed a decreased rate of proteolysis (Figure 2F). Collectively, these results show that structural defects produced by Q655X and R563H increase turnover and/or ER retention of CNG channel subunits.

Characterization of ER stress in 661W cells

Since CNGA3 R563H and Q655X subunits were found to be localization-defective, we hypothesized that these mutations can cause ER stress and induce the UPR. To confirm the suitability of 661W cells as an ER-stress model, we used exogenous agents known to induce ER stress and then characterized the subsequent UPR. Cells were incubated with Tg, Tn or DTT to deplete ER calcium stores, inhibit N-linked glycosylation or prevent disulfide-bond formation respectively. Activation of the three branches of the UPR were monitored by quantifying changes in splicing and transcription of XBP1 mRNA, phosphorylation of PERK, and expression of BiP and CHOP. XBP1 splicing increased within the first hour of each treatment (Figure 3A), indicating a rapid response through the IRE1 branch. Compared with untreated cells, XBP1 transcription increased within 4 h for each treatment, indicating a similarly responsive ATF6 branch (Figure 3B). Transcription of BiP and CHOP also increased within 4 h of treatment (Figures 3C and 3D). Protein indicators of the UPR were analysed by immunoblotting (Figure 3E). PERK phosphorylation was significantly increased (P=0.03) within 30 min, but was reduced (P=0.05) after 6 h of Tg treatment and was unchanged after 6 h of Tn treatment compared with levels in untreated cells (Figure 3F). Levels of BiP increased after 6 h of Tn treatment (P=0.004), but did not respond to Tg treatment. Expression of CHOP increased after 6 h of treatment with Tg (P=0.02) or Tn (P=0.03). These results demonstrate an intact tripartite UPR system in 661W cells and support the use of these markers to monitor ER stress associated with expression of CNG channels.

Pharmacological stress induces the UPR in 661W cells

Figure 3
Pharmacological stress induces the UPR in 661W cells

(AD) RT-qPCR quantification of XBP1 (spliced and total), BiP, CHOP and HPRT (internal standard) mRNA from cells treated with DTT, Tg or Tn, relative to untreated cells. Percentage of spliced XBP1 (A) determined by the percentage of spliced/total mRNA. Fold changes of total XBP1 (B), BiP (C) and CHOP (D). (E and F) Immunoblot quantification of phosphorylated PERK (P-PERK), BiP, CHOP and actin after treatment with Tg or Tn. (E) Representative immunoblot. (F) Levels of P-PERK, BiP and CHOP (relative to actin) after treatments indicated in (E) normalized to levels in untreated cells. Results are means±S.E.M. from three to six independent experiments; *P<0.05.

Figure 3
Pharmacological stress induces the UPR in 661W cells

(AD) RT-qPCR quantification of XBP1 (spliced and total), BiP, CHOP and HPRT (internal standard) mRNA from cells treated with DTT, Tg or Tn, relative to untreated cells. Percentage of spliced XBP1 (A) determined by the percentage of spliced/total mRNA. Fold changes of total XBP1 (B), BiP (C) and CHOP (D). (E and F) Immunoblot quantification of phosphorylated PERK (P-PERK), BiP, CHOP and actin after treatment with Tg or Tn. (E) Representative immunoblot. (F) Levels of P-PERK, BiP and CHOP (relative to actin) after treatments indicated in (E) normalized to levels in untreated cells. Results are means±S.E.M. from three to six independent experiments; *P<0.05.

Localization-defective CNG channels increase ER stress

To test the hypothesis that localization-defective channel subunits induce ER stress, we assessed markers in cells transiently expressing wild-type CNG channel subunits or mutant subunits exhibiting trafficking defects. We also examined CNGB3 F525N, a mutation previously shown to produce hyperactive but localization-competent channels [8]; no significant differences in protein expression levels between CNGB3 wild-type and F525N subunits were detected (normalized to wild-type, 1.00±0.12). Compared with untransfected cells, expression of wild-type CNG channels modestly increased splicing of XBP1 and transcription of BiP and CHOP, but these changes were indistinguishable from expression of a GFP-only control (P=0.2, 0.4 and 0.4 respectively), suggesting a non-specific effect. Compared with wild-type subunits, both R563H and Q655X induced significant increases in splicing of XBP1 and transcription of BiP, whereas F525N expression did not (Figure 4A). CHOP transcription was increased by all three mutations compared with wild-type subunits (Figure 4A). Transcription levels for XBP1 were not significantly different for any mutant compared with wild-type channels (P=0.2, 0.3 and 0.6 for R563H, Q655X and F525N respectively). Finally, we used immunoblotting to examine protein markers of ER stress (Figure 4B). PERK phosphorylation was increased by expression of R563H and Q655X, but not by F525N (Figure 4C). Similarly, increases in BiP and CHOP protein levels were observed only with expression of R563H or Q655X subunits (Figure 4C). Elevation of these markers indicates that expression of localization defective CNG channel subunits increased ER stress in 661W cells.

Localization-defective CNG channels selectively induce the UPR

Figure 4
Localization-defective CNG channels selectively induce the UPR

RT–qPCR and immunoblot quantification of ER-stress markers in cells expressing CNGA3 and CNGB3 subunits. (A) The percentage of spliced XBP1 and fold changes of BiP and CHOP mRNA were normalized to values in cells expressing wild-type channels. Results are means±S.E.M. from three independent experiments; *P<0.05. (B) Representative immunoblot showing expression of P-PERK, BiP, CHOP and actin. (C) Amounts of P-PERK, BiP and CHOP (relative to actin) normalized to levels in cells expressing wild-type channels. Results are means±S.E.M. from four to six independent experiments; *P<0.05.

Figure 4
Localization-defective CNG channels selectively induce the UPR

RT–qPCR and immunoblot quantification of ER-stress markers in cells expressing CNGA3 and CNGB3 subunits. (A) The percentage of spliced XBP1 and fold changes of BiP and CHOP mRNA were normalized to values in cells expressing wild-type channels. Results are means±S.E.M. from three independent experiments; *P<0.05. (B) Representative immunoblot showing expression of P-PERK, BiP, CHOP and actin. (C) Amounts of P-PERK, BiP and CHOP (relative to actin) normalized to levels in cells expressing wild-type channels. Results are means±S.E.M. from four to six independent experiments; *P<0.05.

Chaperones reduce ER stress associated with expression of localization-defective CNG channels

A number of studies have shown that chaperone compounds can reduce ER stress associated with expression of proteins with folding and trafficking defects [34]. We therefore tested representative chaperones for the ability to alleviate ER stress induced by expression of localization-defective CNG channel subunits. Chemical chaperones 4-PBA (5 mM) and TUDCA (1 mM), and the CNG channel ligand analogue CPT-cGMP (1 μM), a putative pharmacological chaperone, were applied to cells expressing CNGA3 wild-type, R563H or Q655X subunits, and the effects on mRNA markers of ER stress were determined. No significant changes were detected in cells expressing wild-type channel subunits (Figures 5A–5C). Splicing of XBP1 mRNA in cells expressing R563H or Q655X was reduced by TUDCA or 4-PBA (Figure 5A); however, CPT-cGMP significantly reduced XBP1 splicing only in cells expressing R563H subunits (Figure 5A). Transcription of BiP displayed the same pattern of change as XBP1 mRNA splicing (Figure 5B). Transcription of CHOP in cells expressing R563H was reduced by TUDCA or CPT-cGMP, and by TUDCA or 4-PBA in cells expressing Q655X (Figure 5C). Transcription of XBP1 was not significantly altered by any treatment (results not shown). Next, we measured the effects of chaperone treatment on protein indicators of ER stress (Figure 6A). No significant changes were detected in cells expressing wild-type channel subunits (Figures 6B and 6C). Application of CPT-cGMP reduced the level of BiP protein in cells expressing R563H (Figure 6B). No other changes in BiP protein levels were significant. The level of CHOP protein in cells expressing R563H or Q655X was reduced by 4-PBA, but not TUDCA treatment, whereas CPT-cGMP treatment significantly reduced CHOP protein only in cells expressing R563H subunits (Figure 6C). No significant changes in phosphorylation of PERK were detected (results not shown). Taken together these results indicate that chaperones can reduce several markers of ER stress in cells expressing localization-defective CNG channel subunits and that CPT-cGMP specifically affected cells expressing channels having a mutation in the cyclic nucleotide-binding domain.

Chemical and pharmacological chaperones reduce mRNA indicators of ER stress induced by localization-defective CNG channels

Figure 5
Chemical and pharmacological chaperones reduce mRNA indicators of ER stress induced by localization-defective CNG channels

RT–qPCR analysis of chaperone-treated cells expressing CNGA3 and CNGB3 subunits. The percentage of spliced XBP1 (A), and fold changes of BiP (B) and CHOP (C) mRNA normalized to values in untreated cells expressing wild-type channels. Results are means±S.E.M. from four to five independent experiments; #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Figure 5
Chemical and pharmacological chaperones reduce mRNA indicators of ER stress induced by localization-defective CNG channels

RT–qPCR analysis of chaperone-treated cells expressing CNGA3 and CNGB3 subunits. The percentage of spliced XBP1 (A), and fold changes of BiP (B) and CHOP (C) mRNA normalized to values in untreated cells expressing wild-type channels. Results are means±S.E.M. from four to five independent experiments; #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Chaperones selectively reduce BiP and CHOP protein levels induced by localization-defective CNG channels

Figure 6
Chaperones selectively reduce BiP and CHOP protein levels induced by localization-defective CNG channels

Immunoblot analysis of chaperone-treated cells expressing CNGA3 and CNGB3 subunits. (A) Representative immunoblot. Amounts of BiP (B) and CHOP (C) protein (relative to actin) normalized to levels in untreated cells expressing wild-type channels. Results are means±S.E.M. from four to five independent experiments. #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Figure 6
Chaperones selectively reduce BiP and CHOP protein levels induced by localization-defective CNG channels

Immunoblot analysis of chaperone-treated cells expressing CNGA3 and CNGB3 subunits. (A) Representative immunoblot. Amounts of BiP (B) and CHOP (C) protein (relative to actin) normalized to levels in untreated cells expressing wild-type channels. Results are means±S.E.M. from four to five independent experiments. #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Chaperones improve maturation and trafficking of localization-defective CNG channels

We next addressed the question of how chaperones reduce ER stress in cells expressing localization-defective CNG channel subunits. To examine the working hypothesis that chaperones alleviate ER stress by assisting maturation and trafficking of proteins, we determined the effect of chaperones on the degradation rate and localization of CNGA3 wild-type, R563H and Q655X subunits. Using cycloheximide treatment to halt protein synthesis (Figure 7A), we found that the stability of Q655X subunits was significantly increased by TUDCA and 4-PBA, but not by CPT-cGMP (Figure 7B). The stability of wild-type and R563H subunits was not significantly changed by any chaperone treatment (Figure 7B). Next, we determined the effect of chaperones on cell-surface localization of CNG channels using cell-impermeant biotinylation (Figure 8A). The surface localization of CNGA3 R563H subunits was restored to wild-type levels by TUDCA and CPT-cGMP (Figure 8B). Surface localization of Q655X subunits was only improved by TUDCA treatment (Figure 8B). Treatment with 4-PBA altered cell morphology; since comparison of surface biotinylation data requires constant surface/volume ratios between samples, the effects of 4-PBA on surface localization of CNGA3 subunits could not be determined. However, surface localization of Q655X and R563H was not significantly reduced compared with wild-type CNGA3 subunits in cells treated with 4-PBA (results not shown). Next, we measured PM localization of CNGA3 subunits using confocal microscopy (Figure 9A). In agreement with cell-surface biotinylation results, PM localization of R563H subunits was increased by TUDCA (P<0.001) and CPT-cGMP (P<0.001) (Figure 9B). Additionally, PM localization of Q655X subunits was significantly increased with TUDCA (P<0.002), but not CPT-cGMP treatment (P=0.6). Finally, we measured ER retention of CNGA3 subunits by co-localization with an ER marker (DsRed2-ER) using confocal microscopy (Figure 9C). Concordant with increased PM localization, ER retention of R563H subunits was reduced by TUDCA (P<0.001) and CPT-cGMP (P<0.001) (Figure 9D). Additionally, ER retention of Q655X subunits was significantly reduced with TUDCA (P<0.001), but not CPT-cGMP treatment (P=0.08). This selectivity of CPT-cGMP rescue for R563H suggests that the mechanism underlying its chaperone effect is mutation-specific. Collectively, these results show that alleviation of ER stress by chaperones is associated with improvement in the maturation and trafficking of localization-defective CNG channel subunits.

Chaperones selectively enhance stability of localization-defective CNGA3 Q655X subunits

Figure 7
Chaperones selectively enhance stability of localization-defective CNGA3 Q655X subunits

Immunoblot analysis of proteins in chaperone-treated cells expressing CNGA3 and CNGB3 subunits after arrest of protein synthesis by cycloheximide. (A) Representative immunoblot. (B) Amount of CNGA3 subunit protein remaining after 4 h relative to the initial amount, normalized to values in untreated cells. Results are means±S.E.M. from four independent experiments; *P<0.05 compared with untreated cells.

Figure 7
Chaperones selectively enhance stability of localization-defective CNGA3 Q655X subunits

Immunoblot analysis of proteins in chaperone-treated cells expressing CNGA3 and CNGB3 subunits after arrest of protein synthesis by cycloheximide. (A) Representative immunoblot. (B) Amount of CNGA3 subunit protein remaining after 4 h relative to the initial amount, normalized to values in untreated cells. Results are means±S.E.M. from four independent experiments; *P<0.05 compared with untreated cells.

Chaperones selectively enhance cell-surface localization of CNGA3 R563H subunits

Figure 8
Chaperones selectively enhance cell-surface localization of CNGA3 R563H subunits

Immunoblot analysis of CNGA3 PM localization using cell-surface biotinylation with avidin-retrieval. (A) Representative immunoblot showing total (1% of input) and surface-localized CNGA3 subunits. (B) PM localization of CNGA3 subunits, displayed relative to untreated cells expressing wild-type channels. Results are means±S.E.M. from three independent experiments; #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Figure 8
Chaperones selectively enhance cell-surface localization of CNGA3 R563H subunits

Immunoblot analysis of CNGA3 PM localization using cell-surface biotinylation with avidin-retrieval. (A) Representative immunoblot showing total (1% of input) and surface-localized CNGA3 subunits. (B) PM localization of CNGA3 subunits, displayed relative to untreated cells expressing wild-type channels. Results are means±S.E.M. from three independent experiments; #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Chaperones selectively enhance plasma-membrane localization and reduce ER retention of localization defective CNGA3 R563H and Q655X subunits

Figure 9
Chaperones selectively enhance plasma-membrane localization and reduce ER retention of localization defective CNGA3 R563H and Q655X subunits

Co-localization of CNGA3 (green) with PM (red) or ER (blue), determined by confocal fluorescence microscopy. (A and C) Representative confocal images. (B and D) Co-localization calculated for regions defined in (A) or (C) using Pearson's product-moment correlation coefficient, normalized to untreated cells expressing wild-type channels. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=4); #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Figure 9
Chaperones selectively enhance plasma-membrane localization and reduce ER retention of localization defective CNGA3 R563H and Q655X subunits

Co-localization of CNGA3 (green) with PM (red) or ER (blue), determined by confocal fluorescence microscopy. (A and C) Representative confocal images. (B and D) Co-localization calculated for regions defined in (A) or (C) using Pearson's product-moment correlation coefficient, normalized to untreated cells expressing wild-type channels. Images are 0.4–0.6 μm optical Z sections, representative of >10 cells each; results are means±S.E.M. (n=4); #P<0.05 compared with untreated cells expressing wild-type channels; *P<0.05 compared with untreated cells.

Chaperones reduce cytotoxicity associated with expression of localization-defective CNG channels

Our working hypothesis, that ER stress plays a role in cell death arising from localization-defective CNG channels, predicts that alleviation of ER stress would promote viability of cells expressing such channels. To test this prediction in our model, we treated cells expressing CNG channels with chaperones and measured the effect on cell viability. Expression of R563H and Q655X subunits increased cell death relative to cells expressing wild-type channels (Figure 10). Chemical chaperones TUDCA and 4-PBA reduced cytotoxicity arising from R563H or Q655X expression, whereas the pharmacological chaperone CPT-cGMP rescued the viability of cells expressing R563H, but not Q655X, subunits (Figure 10). Viability of cells expressing wild-type CNG channels was not significantly affected by chaperone treatment. Taken together these results show that chaperone agents can improve the viability of cells expressing localization-defective CNG channel subunits.

Chaperones improve survival of cells expressing localization-defective CNG channels

Figure 10
Chaperones improve survival of cells expressing localization-defective CNG channels

Quantification of relative cell death, using the LDH release assay. The percentage cell death is shown (relative to vector-only-transfected cells), normalized to untreated cells expressing wild-type channels, indicating the rescue of cell viability with chaperone treatment. Results are means±S.E.M. for three to five independent experiments; #P<0.02 compared with untreated cells expressing wild-type channels; *P<0.01 compared with untreated cells.

Figure 10
Chaperones improve survival of cells expressing localization-defective CNG channels

Quantification of relative cell death, using the LDH release assay. The percentage cell death is shown (relative to vector-only-transfected cells), normalized to untreated cells expressing wild-type channels, indicating the rescue of cell viability with chaperone treatment. Results are means±S.E.M. for three to five independent experiments; #P<0.02 compared with untreated cells expressing wild-type channels; *P<0.01 compared with untreated cells.

DISCUSSION

In the present study, we demonstrate that expression of localization-defective CNG channels bearing disease-associated mutations is associated with induction of the UPR and ER-stress-related cytotoxicity in a photoreceptor-derived cell line. Expression of CNGA3 R563H or Q655X subunits increased markers of ER stress relative to levels associated with expression of wild-type subunits (Figure 4). Increased splicing of XBP1 mRNA indicates signalling through the IRE1 branch of the UPR, and increased phosphorylation of PERK indicates activation of the eponymous UPR pathway. Increased expression of BiP and CHOP denotes activation of multiple branches of the UPR, including ATF6. The adaptive function of the UPR prevents cell death until the magnitude of ER stress reaches a tipping point, after which apoptosis is initiated [26]. Compared with the acute ER-stress response (Figure 3) and rapid cell death induced by chemical agents (results not shown), the milder response to expression of localization-defective CNG channel subunits suggests a mostly adaptive rather than apoptotic ER-stress response over the short assay period examined in the present study [14,26]. These results are consistent with the reported disease progression associated with these and related mutations in humans, which can occur over a span of years [2,3,6], and support our interpretation that expression of localization-defective CNG channel subunits is associated with ER stress.

The results of the present study indicate that induction of the UPR was characteristic of localization-defective, but not localization-competent, channels. In contrast with trafficking-defective CNGA3 subunits, cell death was not increased by expression of wild-type subunits. Similarly, expression of localization-competent CNGB3 F525N did not increase markers of ER stress, with the exception of CHOP transcription (Figure 4). Our previous work has shown that the F525N mutation gives rise to hyperactive CNG channels [8] and promotes calcium-dependent cytotoxicity in 661W cells (C. Liu and M. Varnum, unpublished work). Thus, results of the present study are in agreement with experiments demonstrating that elevated intracellular calcium increases transcription [35] and stability of CHOP mRNA [36] without inducing a co-ordinated ER-stress response. The specificity of the ER-stress response to localization-defective subunits indicates that it is not a general feature of CNG channel expression.

We treated channel-expressing cells with chaperone agents known to improve protein maturation and modulate the UPR in other disease models [22]. Our experiments demonstrate that chaperone treatment reduced ER stress and improved viability in cells expressing trafficking-defective CNG channel subunits. The chemical chaperone 4-PBA increased survival of cells expressing either R563H or Q655X subunits (Figure 10) and reduced markers of ER stress, with the exception of BiP protein levels (Figures 5 and 6). This result suggests an adaptive state in which the endogenous chaperone BiP remains elevated, whereas the pro-apoptotic CHOP returns to unstressed levels. Similarly, the chemical chaperone TUDCA rescued cells from the cytotoxicity associated with R563H or Q655X subunit expression (Figure 10). Although mRNA markers of ER stress were reduced by TUDCA (Figure 5), reductions of BiP and CHOP protein levels were not observed (Figure 6). Although chemical chaperones can stabilize proteins and hydrophobic aggregates indiscriminately, pharmacological chaperones bind and stabilize a specific structural element in a protein (e.g. ligand-binding pocket) in a quasi-native conformation [22]. Notably, the membrane-permeant ligand analogue CPT-cGMP reduced markers of ER stress and cytotoxicity in cells expressing R563H, but had no significant effect on cells expressing Q655X (Figures 5, 6 and 10). The Q655X mutation eliminates a region distal to the CNBD important for subunit assembly [32,37], thus it is not surprising that CPT-cGMP had no effect on cells expressing Q655X subunits. In all cases, improvements in cell viability occurred concomitantly with reduction in ER-stress markers, providing strong evidence to support the hypothesis that ER stress contributes to the deleterious cellular effects of localization-defective CNG channels.

The mechanisms underlying chaperone rescue of cell viability include overt changes in the processing of defective CNG channels. Although progressive degeneration of cone photoreceptors in Cnga3−/− and Cngb3−/− mice demonstrates that CNG channels are necessary for cone photoreceptor function and survival [38,39], in this case it is unlikely that chaperone rescue depends only on restoration of channel activity, since expression of wild-type CNG channels did not improve cell viability (results not shown). Rather, chaperones probably alleviate ER stress by assisting CNG channel-maturation processes such as folding and trafficking. The relative steady-state expression levels of R563H and Q655X remained statistically equivalent to wild-type subunits after chaperone treatment (Figures 7A and 8A, and results not shown). However, Q655X subunits degraded more rapidly than wild-type subunits (Figure 2C), indicating a higher turnover rate. Chemical chaperones decreased the degradation rate of Q655X (Figure 7) without changing relative steady-state levels, thereby reducing the turnover rate of Q655X subunits. Additionally, TUDCA improved surface localization and reduced ER retention of Q655X subunits (Figure 9). Thus decreased processing demand and improved trafficking correlated with reduced ER stress in chaperone-treated cells expressing Q655X subunits. Similarly, TUDCA and CPT-cGMP improved surface localization and reduced ER retention of R563H subunits (Figures 8 and 9), suggesting that improved trafficking alleviates ER stress in cells expressing R563H. Chaperones may support forward trafficking by facilitating protein–protein interactions that promote ER exit (e.g. endogenous chaperones). Chemical chaperones are expected to operate in a general manner, and indeed surface localization of Q655X and R563H was indistinguishable from wild-type channels in cells treated with 4-PBA (results not shown). The pharmacological chaperone CPT-cGMP is expected to aid folding of the CNBD, perhaps improving consequent interactions with proteins promoting forward trafficking. Although trypsin protease sensitivity of R563H was not significantly different from wild-type subunits (Figure 2F), this method may not reveal subtle local structural abnormalities. In agreement with this scenario, Matveev et al. [40] analysed a nearby CNBD mutation and found a local effect on CNBD folding that did not alter proteolytic (trypsin) sensitivity. Collectively, these results suggest that improved CNG channel subunit maturation and trafficking may underlie alleviation of ER stress and improvement of viability in chaperone-treated cells.

Taken together, the results of the present study suggest that localization-defective CNG channel subunits may contribute to the aetiology of retinal disease by inducing ER stress. Genetic and functional defects that give rise to human channelopathies have been extensively studied (reviewed in [41]), and the involvement of ER stress in the context of channel trafficking abnormalities is gaining recognition. It has been proposed that CFRD (CF-related diabetes) is a result of ER-stress-mediated apoptosis of pancreatic β-cells [42], in which the CFTR has been shown to be highly expressed [43]. CFRD is more than 20-fold more commonly associated with mutations that produce localization-defective rather than localization-competent CFTR [44]. Illustrating this, severe pancreatic insufficiency is common in patients homozygous for the ΔF508 mutation [45]. For a HERG channelopathy, expression of the trafficking-defective I593R subunit activates ER-stress pathways that are hypothesized to contribute to the characteristic long QT syndrome, in part by altering calcium-handling physiology [16,46]. Taken together, these studies imply that disease mechanisms arising from trafficking-defective channel proteins are not limited to a loss of channel function, but also may include significant ER-stress-mediated changes in cell physiology and viability.

Cross-talk of the UPR with both the intrinsic and extrinsic apoptotic pathways allows the ER to monitor and integrate other stressors in addition to protein accumulation. Although oxidative stress, perturbed calcium homoeostasis and metabolic overload have been implicated in photoreceptor degeneration [47], appreciation of the connection with ER stress and the UPR is a more recent development. In a study using ERAI (ER-stress-activated indicator) transgenic mice, application of NMDA (N-methyl-D-aspartate) to stimulate glutamatergic activity in the retina caused intracellular calcium overload, nitric oxide production and ER-stress-dependent apoptosis [23]. In the rd1 mouse model of RP, the disabled cGMP phosphodiesterase causes increased CNG channel activity, leading to photoreceptor degeneration reported to arise from disrupted calcium homoeostasis and oxidative stress [48]. Expression of UPR markers in the rd1 mouse retina increased at post-natal day 10 and peaked at post-natal day 12, coinciding with the onset and peak of photoreceptor apoptosis [49]. These observations suggest that ER stress and the UPR are intimately involved in photoreceptor survival and may serve as a central processor of multiple sources of cellular insult.

The UPR has evolved as a protective mechanism, responding to internal and external stresses by returning the ER to homoeostasis or inducing apoptosis in response to acute stress, and by adapting cellular mechanisms to accommodate chronic stress. Alternatively, successful adaptation allows cells to become resistant to future stresses, including those unrelated to the initial insult, in a process known as ‘preconditioning’ [14,50]. If the UPR does serve as a central processor of cellular stressors, then it might in some cases explain the phenotypic heterogeneity of retinal degeneration in individuals with similar genotypes, since environmental factors or genetic background would contribute to the stress load sensed by the ER. Photoreceptors have evolved to withstand a lifetime of physiological demands and environmental exposure. Thus it is conceivable that regulation of their UPR is heavily biased for adaptation and a large accumulation of insults is required to initiate apoptosis. The present study, along with other evidence [23,24,26], suggests involvement of ER stress in photoreceptor degeneration. We envision a model in which trafficking-defective channels induce ER stress and activate the UPR, ultimately leading to apoptosis. On the basis of our model, chaperone therapy might intervene in this process to slow photoreceptor death. In addition, elements of the UPR (e.g. translation attenuation) may complicate gene-replacement strategies designed to restore function in cases involving misfolded and/or mislocalized endogenous channels. Alternatively, gene therapy targeting elements of the UPR itself may prove beneficial. Recognizing the contribution of ER stress and the UPR to the deleterious consequences of CNG channel-trafficking defects may offer new avenues for treatment or prevention of debilitating retinal dystrophies.

Abbreviations

     
  • ACHM

    achromatopsia

  •  
  • ATF6

    activating transcription factor 6

  •  
  • BiP

    immunoglobulin heavy-chain-binding protein

  •  
  • CF

    cystic fibrosis

  •  
  • CFRD

    CF-related diabetes

  •  
  • CFTR

    CF transmembrane conductance regulator

  •  
  • CHOP

    C/EBP (CCAAT/enhancer-binding protein)-homologous protein

  •  
  • CNBD

    cyclic nucleotide-binding domain

  •  
  • CNG

    cyclic nucleotide-gated

  •  
  • CPT-cGMP

    8-(4-chlorophenylthio)-cGMP

  •  
  • DMEM

    Dulbecco's modified Eagle's medium

  •  
  • DPBS

    Dulbecco's PBS

  •  
  • DTT

    dithiothreitol

  •  
  • EGFP

    enhanced green fluorescent protein

  •  
  • ER

    endoplasmic reticulum

  •  
  • FBS

    fetal bovine serum

  •  
  • GFP

    green fluorescent protein

  •  
  • HERG

    human ether-a-go-go-related gene

  •  
  • HPRT

    hypoxanthine–guanine phosphoribosyltransferase

  •  
  • IRE1

    inositol-requiring enzyme 1

  •  
  • LDH

    lactate dehydrogenase

  •  
  • MD

    macular degeneration

  •  
  • NGS

    normal goat serum

  •  
  • 4-PBA

    sodium 4-phenylbutyrate

  •  
  • PCD

    progressive cone dystrophy

  •  
  • PERK

    PKR (double-stranded-RNA-dependent protein kinase)-like ER kinase

  •  
  • PM

    plasma membrane

  •  
  • RP

    retinitis pigmentosa

  •  
  • RT–qPCR

    reverse transcription–quantitative PCR

  •  
  • TBS

    Tris-buffered saline

  •  
  • Tg

    thapsigargin

  •  
  • Tn

    tunicamycin

  •  
  • TUDCA

    tauroursodeoxycholate sodium salt

  •  
  • UPR

    unfolded protein response

  •  
  • WGA

    wheat germ agglutinin

  •  
  • XBP1

    X-box-binding protein 1

AUTHOR CONTRIBUTION

Michael Varnum conceived and supervised the project. Deborah Duricka, Lane Brown and Michael Varnum designed the experiments. Experiments were carried out by Deborah Duricka. Deborah Duricka, Lane Brown and Michael Varnum analysed the data. Deborah Duricka, Lane Brown and Michael Varnum wrote the paper.

We thank E. Rich for expert technical support, and J. Harding and L. Orfe for critical comments on the paper prior to submission. We also thank Professor M.R. Al-Ubaidi for providing the 661W cell line, and Professor K.-W. Yau (Johns Hopkins University, Baltimore, MD, U.S.A.) for sharing the cDNA clone for human CNGA3.

FUNDING

This work was supported by the National Institutes of Health/National Eye Institute [grant number EY12836 (to M.D.V.) and EY19907 (to R.L.B.)], the Adler Foundation and the Autzen Endowment (to M.D.V.), and by a Poncin Award (to D.L.D.).

References

References
1
Biel
M.
Michalakis
S.
Function and dysfunction of CNG channels: insights from channelopathies and mouse models
Mol. Neurobiol.
2007
, vol. 
35
 (pg. 
266
-
277
)
2
Khan
N. W.
Wissinger
B.
Kohl
S.
Sieving
P. A.
CNGB3 achromatopsia with progressive loss of residual cone function and impaired rod-mediated function
Invest. Ophthalmol. Visual Sci.
2007
, vol. 
48
 (pg. 
3864
-
3871
)
3
Thiadens
A. A.
Slingerland
N. W.
Roosing
S.
van Schooneveld
M. J.
van Lith-Verhoeven
J. J.
van Moll-Ramirez
N.
van den Born
L. I.
Hoyng
C. B.
Cremers
F. P.
Klaver
C. C.
Genetic etiology and clinical consequences of complete and incomplete achromatopsia
Ophthalmology
2009
, vol. 
116
 (pg. 
1984
-
1989.e1
)
4
Nishiguchi
K. M.
Sandberg
M. A.
Gorji
N.
Berson
E. L.
Dryja
T. P.
Cone cGMP-gated channel mutations and clinical findings in patients with achromatopsia, macular degeneration, and other hereditary cone diseases
Hum. Mutat.
2005
, vol. 
25
 (pg. 
248
-
258
)
5
Kohl
S.
Baumann
B.
Broghammer
M.
Jagle
H.
Sieving
P.
Kellner
U.
Spegal
R.
Anastasi
M.
Zrenner
E.
Sharpe
L. T.
Wissinger
B.
Mutations in the CNGB3 gene encoding the β-subunit of the cone photoreceptor cGMP-gated channel are responsible for achromatopsia (ACHM3) linked to chromosome 8q21
Hum. Mol. Genet.
2000
, vol. 
9
 (pg. 
2107
-
2116
)
6
Wissinger
B.
Gamer
D.
Jagle
H.
Giorda
R.
Marx
T.
Mayer
S.
Tippmann
S.
Broghammer
M.
Jurklies
B.
Rosenberg
T.
, et al. 
CNGA3 mutations in hereditary cone photoreceptor disorders
Am. J. Hum. Genet.
2001
, vol. 
69
 (pg. 
722
-
737
)
7
Liu
C.
Varnum
M. D.
Functional consequences of progressive cone dystrophy-associated mutations in the human cone photoreceptor cyclic nucleotide-gated channel CNGA3 subunit
Am. J. Physiol.
2005
, vol. 
289
 (pg. 
C187
-
C198
)
8
Bright
S. R.
Brown
T. E.
Varnum
M. D.
Disease-associated mutations in CNGB3 produce gain of function alterations in cone cyclic nucleotide-gated channels
Mol. Vision
2005
, vol. 
11
 (pg. 
1141
-
1150
)
9
Peng
C.
Rich
E. D.
Varnum
M. D.
Achromatopsia-associated mutation in the human cone photoreceptor cyclic nucleotide-gated channel CNGB3 subunit alters the ligand sensitivity and pore properties of heteromeric channels
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
34533
-
34540
)
10
Muraki-Oda
S.
Toyoda
F.
Okada
A.
Tanabe
S.
Yamade
S.
Ueyama
H.
Matsuura
H.
Ohji
M.
Functional analysis of rod monochromacy-associated missense mutations in the CNGA3 subunit of the cone photoreceptor cGMP-gated channel
Biochem. Biophys. Res. Commun.
2007
, vol. 
362
 (pg. 
88
-
93
)
11
Reuter
P.
Koeppen
K.
Ladewig
T.
Kohl
S.
Baumann
B.
Wissinger
B.
Mutations in CNGA3 impair trafficking or function of cone cyclic nucleotide-gated channels, resulting in achromatopsia
Hum. Mutat.
2008
, vol. 
29
 (pg. 
1228
-
1236
)
12
Patel
K. A.
Bartoli
K. M.
Fandino
R. A.
Ngatchou
A. N.
Woch
G.
Carey
J.
Tanaka
J. C.
Transmembrane S1 mutations in CNGA3 from achromatopsia 2 patients cause loss of function and impaired cellular trafficking of the cone CNG channel
Invest. Ophthalmol. Visual Sci.
2005
, vol. 
46
 (pg. 
2282
-
2290
)
13
Schroder
M.
Kaufman
R.
ER stress and the unfolded protein response
Mutat. Res.
2005
, vol. 
569
 (pg. 
29
-
63
)
14
Rutkowski
D. T.
Arnold
S. M.
Miller
C. N.
Wu
J.
Li
J.
Gunnison
K. M.
Mori
K.
Sadighi Akha
A. A.
Raden
D.
Kaufman
R. J.
Adaptation to ER stress is mediated by differential stabilities of pro-survival and pro-apoptotic mRNAs and proteins
PLoS Biol.
2006
, vol. 
4
 pg. 
e374
 
15
Madeo
F.
Kroemer
G.
Intricate links between ER stress and apoptosis
Mol. Cell
2009
, vol. 
33
 (pg. 
669
-
670
)
16
Keller
S. H.
Platoshyn
O.
Yuan
J. X.
Long QT syndrome-associated I593R mutation in HERG potassium channel activates ER stress pathways
Cell. Biochem. Biophys.
2005
, vol. 
43
 (pg. 
365
-
377
)
17
Kerbiriou
M.
Le Drévo
M.
Férec
C.
Trouvé
P.
Coupling cystic fibrosis to endoplasmic reticulum stress: differential role of Grp78 and ATF6
Biochim. Biophys. Acta
2007
, vol. 
1772
 (pg. 
1236
-
1249
)
18
Singh
O. V.
Pollard
H. B.
Zeitlin
P. L.
Chemical rescue of ΔF508-CFTR mimics genetic repair in cystic fibrosis bronchial epithelial cells
Mol. Cell. Proteomics
2008
, vol. 
7
 (pg. 
1099
-
1110
)
19
Zhou
Z.
Gong
Q.
January
C. T.
Correction of defective protein trafficking of a mutant HERG potassium channel in human long QT syndrome. Pharmacological and temperature effects
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
31123
-
31126
)
20
Ficker
E.
Obejero-Paz
C. A.
Zhao
S.
Brown
A. M.
The binding site for channel blockers that rescue misprocessed human long QT syndrome type 2 ether-a-gogo-related gene (HERG) mutations
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
4989
-
4998
)
21
Noorwez
S. M.
Malhotra
R.
McDowell
J. H.
Smith
K. A.
Krebs
M. P.
Kaushal
S.
Retinoids assist the cellular folding of the autosomal dominant retinitis pigmentosa opsin mutant P23H
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
16278
-
16284
)
22
Mendes
H. F.
Cheetham
M. E.
Pharmacological manipulation of gain-offunction and dominant-negative mechanisms in rhodopsin retinitis pigmentosa
Hum. Mol. Genet.
2008
, vol. 
17
 (pg. 
3043
-
3054
)
23
Shimazawa
M.
Inokuchi
Y.
Ito
Y.
Murata
H.
Aihara
M.
Miura
M.
Araie
M.
Hara
H.
Involvement of ER stress in retinal cell death
Mol. Vision
2007
, vol. 
13
 (pg. 
578
-
587
)
24
Inokuchi
Y.
Nakajima
Y.
Shimazawa
M.
Kurita
T.
Kubo
M.
Saito
A.
Sajiki
H.
Kudo
T.
Aihara
M.
Imaizumi
K.
, et al. 
Effect of an inducer of BiP, a molecular chaperone, on endoplasmic reticulum (ER) stress-induced retinal cell death
Invest. Ophthalmol. Visual Sci.
2009
, vol. 
50
 (pg. 
334
-
344
)
25
Gorbatyuk
M. S.
Knox
T.
LaVail
M. M.
Gorbatyuk
O. S.
Noorwez
S. M.
Hauswirth
W. W.
Lin
J. H.
Muzyczka
N.
Lewin
A. S.
Restoration of visual function in P23H rhodopsin transgenic rats by gene delivery of BiP/Grp78
Proc. Natl. Acad. Sci. U.S.A.
2010
, vol. 
107
 (pg. 
5961
-
5966
)
26
Lin
J. H.
Li
H.
Yasumura
D.
Cohen
H. R.
Zhang
C.
Panning
B.
Shokat
K. M.
Lavail
M. M.
Walter
P.
IRE1 signaling affects cell fate during the unfolded protein response
Science
2007
, vol. 
318
 (pg. 
944
-
949
)
27
Surgucheva
I.
Ninkina
N.
Buchman
V. L.
Grasing
K.
Surguchov
A.
Protein aggregation in retinal cells and approaches to cell protection
Cell. Mol. Neurobiol.
2005
, vol. 
25
 (pg. 
1051
-
1066
)
28
Peng
C.
Rich
E. D.
Varnum
M. D.
Subunit configuration of heteromeric cone cyclic nucleotide-gated channels
Neuron
2004
, vol. 
42
 (pg. 
401
-
410
)
29
Pfaffl
M. W.
A new mathematical model for relative quantification in real-time RT-PCR
Nucleic Acids Res.
2001
, vol. 
29
 pg. 
e45
 
30
de Kok
J. B.
Roelofs
R. W.
Giesendorf
B. A.
Pennings
J. L.
Waas
E. T.
Feuth
T.
Swinkels
D. W.
Span
P. N.
Normalization of gene expression measurements in tumor tissues: comparison of 13 endogenous control genes
Lab. Invest.
2005
, vol. 
85
 (pg. 
154
-
159
)
31
Fitzgerald
J. B.
Malykhina
A. P.
Al-Ubaidi
M. R.
Ding
X. Q.
Functional expression of cone cyclic nucleotide-gated channel in cone photoreceptor-derived 661W cells
Adv. Exp. Med. Biol.
2008
, vol. 
613
 (pg. 
327
-
334
)
32
Zhong
H.
Lai
J.
Yau
K. W.
Selective heteromeric assembly of cyclic nucleotide-gated channels
Proc. Natl. Acad. Sci. U.S.A.
2003
, vol. 
100
 (pg. 
5509
-
5513
)
33
Tan
E.
Ding
X. Q.
Saadi
A.
Agarwal
N.
Naash
M. I.
Al-Ubaidi
M. R.
Expression of cone-photoreceptor-specific antigens in a cell line derived from retinal tumors in transgenic mice
Invest. Ophthalmol. Visual Sci.
2004
, vol. 
45
 (pg. 
764
-
768
)
34
Engin
F.
Hotamisligil
G. S.
Restoring endoplasmic reticulum function by chemical chaperones: an emerging therapeutic approach for metabolic diseases
Diabetes, Obes. Metab.
2010
, vol. 
12
 
Suppl. 2
(pg. 
108
-
115
)
35
Copanaki
E.
Schurmann
T.
Eckert
A.
Leuner
K.
Muller
W. E.
Prehn
J. H.
Kogel
D.
The amyloid precursor protein potentiates CHOP induction and cell death in response to ER Ca2+ depletion
Biochim. Biophys. Acta
2007
, vol. 
1773
 (pg. 
157
-
165
)
36
Bartlett
J. D.
Luethy
J. D.
Carlson
S. G.
Sollott
S. J.
Holbrook
N. J.
Calcium ionophore A23187 induces expression of the growth arrest and DNA damage inducible CCAAT/enhancer-binding protein (C/EBP)-related gene, gadd153
J. Biol. Chem.
1992
, vol. 
267
 (pg. 
20465
-
20470
)
37
Zhong
H.
Molday
L. L.
Molday
R. S.
Yau
K. W.
The heteromeric cyclic nucleotide-gated channel adopts a 3A:1B stoichiometry
Nature
2002
, vol. 
420
 (pg. 
193
-
198
)
38
Michalakis
S.
Geiger
H.
Haverkamp
S.
Hofmann
F.
Gerstner
A.
Biel
M.
Impaired opsin targeting and cone photoreceptor migration in the retina of mice lacking the cyclic nucleotide-gated channel CNGA3
Invest. Ophthalmol. Visual Sci.
2005
, vol. 
46
 (pg. 
1516
-
1524
)
39
Ding
X. Q.
Harry
C. S.
Umino
Y.
Matveev
A. V.
Fliesler
S. J.
Barlow
R. B.
Impaired cone function and cone degeneration resulting from CNGB3 deficiency: down-regulation of CNGA3 biosynthesis as a potential mechanism
Hum. Mol. Genet.
2009
, vol. 
18
 (pg. 
4770
-
4780
)
40
Matveev
A. V.
Fitzgerald
J. B.
Xu
J.
Malykhina
A. P.
Rodgers
K. K.
Ding
X. Q.
The disease-causing mutations in the carboxyl terminus of the cone cyclic nucleotide-gated channel CNGA3 subunit alter the local secondary structure and interfere with the channel active conformational change
Biochemistry
2010
, vol. 
49
 (pg. 
1628
-
1639
)
41
Ashcroft
F. M.
From molecule to malady
Nature
2006
, vol. 
440
 (pg. 
440
-
447
)
42
Ali
B. R.
Is cystic fibrosis-related diabetes an apoptotic consequence of ER stress in pancreatic cells?
Med. Hypotheses
2009
, vol. 
72
 (pg. 
55
-
57
)
43
Polychronakos
C.
Early onset diabetes mellitus. Tip or iceberg?
Pediatr. Diabetes
2004
, vol. 
5
 (pg. 
171
-
173
)
44
Koch
C.
Cuppens
H.
Rainisio
M.
Madessani
U.
Harms
H.
Hodson
M.
Mastella
G.
Navarro
J.
Strandvik
B.
McKenzie
S.
European Epidemiologic Registry of Cystic Fibrosis (ERCF): comparison of major disease manifestations between patients with different classes of mutations
Pediatr. Pulmonol.
2001
, vol. 
31
 (pg. 
1
-
12
)
45
Marshall
B. C.
Butler
S. M.
Stoddard
M.
Moran
A. M.
Liou
T. G.
Morgan
W. J.
Epidemiology of cystic fibrosis-related diabetes
J. Pediatr.
2005
, vol. 
146
 (pg. 
681
-
687
)
46
Thuerauf
D. J.
Hoover
H.
Meller
J.
Hernandez
J.
Su
L.
Andrews
C.
Dillmann
W. H.
McDonough
P. M.
Glembotski
C. C.
Sarco/endoplasmic reticulum calcium ATPase-2 expression is regulated by ATF6 during the endoplasmic reticulum stress response: intracellular signaling of calcium stress in a cardiac myocyte model system
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
48309
-
48317
)
47
Pierce
EA.
Pathways to photoreceptor cell death in inherited retinal degenerations
BioEssays
2001
, vol. 
23
 (pg. 
605
-
618
)
48
Sharma
A. K.
Rohrer
B.
Sustained elevation of intracellular cGMP causes oxidative stress triggering calpain-mediated apoptosis in photoreceptor degeneration
Curr. Eye Res.
2007
, vol. 
32
 (pg. 
259
-
269
)
49
Yang
L. P.
Wu
L. M.
Guo
X. J.
Tso
M. O.
Activation of endoplasmic reticulum stress in degenerating photoreceptors of the rd1 mouse
Invest. Ophthalmol. Visual Sci.
2007
, vol. 
48
 (pg. 
5191
-
5198
)
50
Mendes
C. S.
Levet
C.
Chatelain
G.
Dourlen
P.
Fouillet
A.
Dichtel-Danjoy
M. L.
Gambis
A.
Ryoo
H. D.
Steller
H.
Mollereau
B.
ER stress protects from retinal degeneration
EMBO J.
2009
, vol. 
28
 (pg. 
1296
-
1307
)