A Naturally Occurring Mouse Model of Achromatopsia: Characterization of the Mutation in Cone Transducin and Subsequent Retinal Phenotype

Abstract

Purpose.: This work investigates a novel, naturally occurring mouse model of achromatopsia. The specific missense mutation within the Gnat2 gene was identified and the subsequent retinal phenotype characterized.

Methods.: The Gnat2 sequence was amplified using PCR from BALB/c and Gnat2^c.518A>G retinae and the product sequenced. Retinal function was assessed at 3, 6, 9, and 12 months using the electroretinogram. Transducin and opsin expression were assessed at 3 and 12 months using immunohistochemistry and quantitative PCR. Retinal remodeling and Müller cell gliosis were investigated using immunocytochemistry.

Results.: An A to G missense mutation at position 518 of the Gnat2 gene was identified that resulted in an aspartic acid to glycine substitution. Gnat2^c.518A>G animals showed no cone response, while the rod response was normal except for a decrease in the photoreceptor response at 12 months (a-wave, −14%). Gnat2^c.518A>G retinal sections showed no transducin immunolabeling; however, protein was detected via Western blot. Gnat2 gene expression was only decreased at 12 months of age (−27%). There was reduced cone number at 12 months (−27%) and M-opsin showed evidence of mislocalization. Displaced photoreceptor terminals and altered horizontal cell, cone/rod bipolar cell morphology were evident at 3 months, becoming more extensive at 12 months with the emergence of Müller cell gliosis.

Conclusions.: The Gnat2^c.518A>G mouse contains a missense mutation that results in no cone function due to a misfolding of transducin. Cone photoreceptors also show signs of opsin mislocalization, retinal remodeling and degeneration. This naturally occurring model shows all the hallmark signs of achromatopsia.

Introduction

Complete achromatopsia is a heterogeneous, recessively inherited visual disorder that arises due to a specific loss of cone photoreceptor function. Clinically this disorder, which has a prevalence of 1/30,000, ^1,2 is characterized by loss of color vision, poor visual acuity, photophobia, pendular nystagmus, and absent or severely decreased cone electroretinogram (ERG) recordings. ^3,4 In contrast, affected individuals generally exhibit normal or slightly decreased rod photoreceptor function. ⁵ Achromatopsia has a heterogeneous retinal phenotype, and while retinal structure usually appears normal via funduscopy, there are reports of pigment changes, altered cone inner/outer segment structure, and even complete cone photoreceptor loss. ^5,6  

At present, mutations leading to achromatopsia have been mapped to four different genes, all of which are involved in the cone phototransduction process. Almost three quarters of the achromatopsia cases arise due to mutations in either the α or β subunits of the cone cyclic nucleotide-gated channel (CNGA3 and CNGB3, respectively), which is responsible for photoreceptor hyperpolarization. ^7,8 Causative mutations have also been mapped to the α subunit of cone transducin (GNAT2) ^9,10 and cone cyclic nucleotide phosphodiesterase (PDE6C), ¹¹ both of which play roles in the early phototransduction cascade. Multiple disease-causing mutations have been described for each of these four genes, with at least 10 different GNAT2 mutations alone being identified. ^9,12−14  

Currently there are mouse models for the four proteins identified in the human linkage studies, ^11,15−18 with two larger animal models also described (canine CNGB^−/− and sheep CNGA3^−/− ). ^19,20 Most work has investigated the CngA3^−/− and CngB3^−/− mice, with no appreciable difference in the respective phenotypes. Generally they are characterized by no or severely decreased cone ERG response, mislocalized or decreased cone photoreceptor proteins, disrupted cone outer segment structure, cone cell loss, and subsequent retinal remodeling. ^15,17,21,22 Recent work also shows reduced rod b-waves in the CngA3^−/− by 1 month of age, with subsequent rod a-wave deficits and cell loss by 12 months. ²³ The studies characterizing the Gnat2^cpfl3 and cpfl1 (Pde6C) mice generally reflect the cone-related phenotype; however, cone cell loss and rod pathway involvement are unclear. ^11,16  

Since achromatopsia has a relatively simple etiology arising from a single mutant protein, its treatment via gene delivery strategies has great promise. Several reports have investigated such treatment options in mouse ²⁴⁻²⁶ and canine ²⁷ models and show that delivery of the correct protein results in restoration of cone ERG, proper cone photoreceptor protein trafficking, downstream neuronal signaling, and improved visual guided behavior. Despite these positive results, the restorative potential has been variable, with the use of different promoter constructs and model systems providing cone ERG improvements ranging from 10% (canine CNGB3^−/− model) to 90% (CngB3^−/− murine model). ^26,27  

The current study characterizes a novel, spontaneously occurring mutation in the murine Gnat2 gene that results in an achromatopsia-like phenotype. The specific single base-pair substitution was identified in exon 5, and cone transducin protein expression was investigated. Retinal function was assessed with age, and retinal remodeling was probed by using immunohistochemistry. This new mouse model may prove useful in the investigation of development of gene-related treatments for achromatopsia. 

Methods

Animals

The initial strain (Cx3cr1^gfp/gfp/BALB/c) was sourced from Paul McMenamin and was originally derived from Jung et al. ²⁸ These animals had the coding sequence for enhanced green fluorescent protein (eGFP) placed within exon 1 of the fractalkine receptor, Cx3cr1. Our preliminary analysis showed that, in addition to the eGFP transgene, this particular strain contained an additional mutation in the cone transducin gene (Gnat2). Subsequent backcrossing of the Cx3cr1^gfp/gfp/BALB/c with BALB/c (wild-type [WT]) animals (Animal Resources Centre, Canning Vale, Western Australia, Australia) enabled the separation of the Cx3cr1null and Gnat2 mutations and led to the propagation of the strain designated Gnat2^c.518A>G. Age-matched BALB/c control (WT; Animal Resources Centre) and Gnat2^c.518A>G animals were housed under the same conditions for 3 to 12 months and were exposed to light levels of <35 lux. Animals were maintained on a 12-hour light/dark cycle with access to food and water ad libitum. All experiments followed the National Health and Medical Research Council of Australia guidelines and were in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. 

Electroretinogram

Retinal function was assessed every 3 months (3, 6, 9, and 12 months) in Gnat2^c.518A>G animals and age-matched BALB/c controls using ERG. Animals (n > 9 each group) were dark adapted overnight, anesthetised (ketamine:xylazine, 67:13 mg/kg), the corneal reflex was further anesthetised (0.5%; Alcaine, Alcon Laboratories; Frenchs Forest, New South Wales, Australia) and pupils were dilated with tropicamide (0.5%; Mydriacyl; Alcon Laboratories). Animals were placed on heating pad to maintain body temperature. 

A custom-made AgCl recording electrode was placed centrally on the cornea with the reference electrode placed orally. To elicit the ERG response, a full field flash of 2.1 log cd.s/m² was generated by a photography flash (Nikon SB900; Nikon, Lidcombe, New South Wales, Australia) and delivered via a custom-made Ganzfeld. Specifically, two flashes were delivered with a 0.8-second interstimulus interval, to elicit responses from the rod and cone pathways (mixed response) and the cone pathway (cone response), respectively. The cone response was digitally subtracted from the mixed response to generate the rod response. ^29,30 Coordination of ERG stimulation and recording of electrical responses was completed using Scope v3.6.9 software and the responses were filtered for 60-Hz noise, amplified, and digitized at 10 kHz over a 200-ms epoch (gain × 5000; −3 dB at 1 Hz and 1 kHz; ADInstruments, Bella Vista, New South Wales, Australia). 

ERG component analysis was completed using previously published equations and techniques. ^31,32 Rod photoreceptor responses (rod a-wave) were analyzed by using a modified PIII model, and the amplitude of the PIII response (PIII Rmax in microvolts) and the sensitivity (S in m² cd⁻¹ s⁻³) were derived. The rod postphotoreceptoral function (rod b-wave) was isolated by subtraction of the rod PIII from the raw rod waveform and then fitted using an inverted γ function to generate the rod PII from which the amplitude (rod PII Rmax in microvolts) and time to peak (implicit time in milliseconds) could be derived. The oscillatory potentials (OPs) were isolated by subtracting the fitted PIIs from the raw waveforms. In the case of the rod ERG, OP2, OP3, and OP4 were analyzed to assess their amplitude (microvolts) and their implicit time (milliseconds). Due to the small number of cone photoreceptors in the mouse, the cone photoreceptor response (cone a-wave) was too small to be analyzed. However, the cone postreceptoral response (cone b-wave) could be assessed and the cone PII was analyzed by fitting an inverted γ function to the raw cone waveform. From the cone PII fit, the amplitude of the cone PII response (cone PII Rmax in microvolts) and the time to peak (implicit time in milliseconds) were determined. 

Tissue Isolation

Animals were anesthetised with a combination of ketamine and xylazine (67:13 mg/kg) and killed via cervical dislocation. Eyes were enucleated, an incision was made posterior to the limbus, and the cornea and lens were removed. For immunohistochemistry, the posterior eye cup was placed in 4% paraformaldehyde (see following text for details). For molecular analysis, the retina was dissected from the posterior eye cup, snap frozen in liquid nitrogen, and stored at −80°C until use. 

Genotyping and Quantitative PCR

Genomic DNA was isolated from BALB/c and Gnat2^c.518A>G tail samples and mouse-specific Gnat2 primers (forward, 5′-TTTAAGAGCCAGAGGGAACG-3′; reverse, 5′-AGGATGGGCTGATTTGACAC-3′) were used to amplify a 1214-bp fragment incorporating the full Gnat2 coding sequence. The product was purified (Qiaquick; Qiagen, Valencia, CA) and sequenced (Australian Genome Research Facility, Melbourne, Australia). For routine genotyping, primers designed to flank exon 5 of murine Gnat2 (forward, 5′-ACCGATGCCACCTTCTTTTT-3′; reverse, 5′-TGCTGTGAGACCTGAGATGC-3′) were used to amplify a 186-bp product that was digested with the restriction endonuclease Cac8I (New England Biolabs, Ipswich, MA) to identify the WT (186 bp), heterozygous (186, 81, and 105 bp), and mutant (81 and 105 bp) genotypes. 

Total RNA was isolated from 3- and 12-month retinal tissue from BALB/c and Gnat2^c.518A>G animals (each n = 9) using commercial spin columns (RNeasy; Qiagen) incorporating an on-column DNase I digest to remove genomic contamination. Reverse transcription reactions were performed on 500 ng total RNA using random hexamer primers (Tetro; Bioline, London, UK) and subsequently diluted to 5 ng/μL. For quantitative real-time PCR (qPCR), the expression of several phototransduction-related ¹⁷ and opsin genes were assessed relative to the housekeeping genes hypoxanthine guanine phophoribosyl transferase (Hprt) and glyceraldehyde-3-phosphate dehydrogenase (Gapdh, see the Table for primer sequences). External standards were produced using the same primer pairs described in the Table that incorporated T7 promoter (5′-taatacgactcactataggg-3′) and polyT₁₅ sequences at the 5′ end of each forward and reverse primer, respectively. The amplified standards were transcribed into copy RNA (Megascript T7 High Yield Transcription kit; Ambion Inc., Austin, TX) and dilutions combined with yeast t-RNA (500 ng; Invitrogen, Carlsbad, CA) to reflect the retinal total RNA amount (500 ng) used in the reverse transcription reaction. The RNA standards were reverse transcribed with the retinal RNA samples to standardize efficiency. 

Table.

 

Mouse-Specific Oligonucleotide Primers Used in qPCR*

Gene	Sequence	Forward Primer	Reverse Primer	Product Size, bp
Opn1mw	NM_008106.2	GCCCAGACGTGTTCAGCG	GACCATCACCACCACCAT	212
Opn1sw	NM_007538.3	CAGCCTTCATGGGATTTG	GTGCATGCTTGGAGTTGA	339
Rho	NM_145383.1	AGCAGCAGGAGTCAGCCACC	CCGAAGTTGGAGCCCTGGTG	145
Gnat2	NM_008141.2	AGCTCAATGACTCGGCATCT	ACTTCGTCATCCTCCACCAG	280
CngA3	NM_009918.1	GAGGTGAAGGTCCTGCTCTG	GTTATCTCACCAGCCCCGTA	243
CngB3	NM_013927.2	GGCAAAAGGAGGAGGAAATC	TTAAGCATTCGGGGTGTCTC	243
Hprt	NM_013556.2	CCTAAGATGAGCGCAAGTTGAA	CCACAGGACTAGAACACCTGCTAA	86
Gapdh	NM_008084.2	TGTGTCCGTCGTGGATCTGA	TTGCTGTTGAAGTCGCAGGAG	150

*  Mouse-specific primers were designed relative to the respective published sequences as indicated, and the sequences are shown 5′ to 3′ with the amplified product sizes also included. Gnat2, CngA3, and CngB3 primers were taken from Ding et al. ¹⁷ For the external standards the primers contained a T7 promoter sequence at the 5′ end of the forward primers, while a poly T15 sequence was included at the 5′ end of the reverse primers.

Real-time PCR was performed on the Rotorgene V3000 (Corbett Research, Chadstone Centre, Victoria, Australia) using a commercial reaction mixture incorporating SYBR green (SensiFast; Bioline). Respective four-point standard curves were included in every run, and standards and samples were amplified in triplicate. Each primer set yielded only one product of the correct size, and negative controls were included in every run. Absolute gene copy number was calculated with reference to the standard curve (Rotorgene V6.1 software; Corbett Research) and expressed relative to Hprt and Gapdh. 

Immunohistochemistry

Posterior eyecups from 3- and 12-month-old animals (BALB/c and Gnat2^c.518A>G, n = 6) were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4 (PB) for 30 minutes, rinsed in PB, and cryoprotected (10%, 20%, and 30% sucrose in PB). The eye cups were subsequently embedded in Tissue-Tek (OCT; Bayer Diagnostics, Puteaux, France), cryosectioned at 14 μm at −20°C (Microm, Walldorf, Germany), collected on poly-L-lysine-coated slides, and stored at −20°C until use. 

For immunohistochemistry, sections were blocked (10% normal goat serum [Chemicon, Temecula, CA], 1% bovine serum albumin [Sigma, St. Louis, MO], 0.5% Triton X-100 in PB) for 1 hour and incubated overnight at room temperature with the respective primary antibodies described later (primary antibodies diluted in 3% normal goat serum, 1% bovine serum albumin, and 0.5% Triton X-100 in PB). Sections were rinsed in PB, incubated with the respective secondary antibodies for 1 hour, and then mounted (Mowiol 488; Polysciences Inc., Warrington, PA). Sections were imaged (LSM confocal microscope; Zeiss, Göttingen, Germany) using ×20 air or ×40 oil objectives, and images were subsequently adjusted for contrast and brightness using Adobe Photoshop CS4 (Adobe Systems, San Jose, CA). 

In order to assess cone transducin (Gnat2) expression and cone photoreceptor number, sections were labeled with an antibody to Gnat2 (rabbit antibovine G_αt2, 1:100, 1:1000, and 1:10,000; Santa Cruz Biotechnology Inc., Santa Cruz, CA) and subsequently incubated with a goat antirabbit Alexa Fluor 594 (1:500; Molecular Probes, Inc., Eugene, OR) and FITC-conjugated peanut agglutinin (PNA, 1:250; Vector Laboratories, Burlingame, CA) to label cone cell bodies. All nuclei were counterstained with DAPI (0.2 μg/mL; Invitrogen), and cone photoreceptor number was counted and expressed per millimeter of retina. To investigate photoreceptor opsin expression, sections were labeled with antibodies to short wavelength opsin (S-opsin; 1:100,000; rabbit antihuman; Chemicon), medium/long wavelength opsin (M-opsin; 1:500; rabbit antihuman; Chemicon), or rhodopsin (1:1000; mouse antirat; Chemicon) followed by incubation with goat antirabbit or goat antimouse Alexa Fluor 488 (1:500; Molecular Probes). 

Horizontal cell remodeling was assessed in Gnat2^c.518A>G and BALB/c retinae by colabeling sections with antibodies to calbindin (1:2000; Swant, Bellinzona, Switzerland) and vesicular glutamate transporter 1 (VGlut1; 1:1000, guinea pig antirat; Chemicon). Remodeling of rod bipolar cells was investigated using antibodies to protein kinase Cα (PKCα, 1:400; mouse antibovine; Sigma) and VGlut1, while cone bipolar cells were colabeled with ZNP-1 (1:2000; mouse antizebrafish; Zebrafish International Resource Centre, University of Oregon, Eugene, OR) and VGlut1. After secondary antibody incubation (1:500 of goat antiguinea pig Alexa Fluor 594 and goat antimouse Alexa Fluor 488; Molecular Probes) sections were processed and imaged as already described. 

Müller cell gliosis was determined by colabeling sections with antibodies to glial fibrillary acidic protein (GFAP, 1:25,000; rabbit antibovine; Dako, Glostrup, Denmark) and glutamine synthetase (1:1000; mouse antisheep; Chemicon). After the addition of secondary antibodies (1:500, goat antimouse Alexa Fluor 488 and goat antirabbit Alexa Fluor 594; Molecular Probes), sections were processed and imaged as already described. 

Western Blot

Western blots were performed as previously described. ³³ Briefly, Gnat2^c.518A>G and BALB/c retinae were isolated from 3-month-old animals, homogenized (40 mM HEPES, 320 mM sucrose, protease inhibitors; Complete mini protease inhibitors; Roche, Mannheim, Germany) and protein concentration estimated (BCA; Thermo Fisher Scientific, Rockford, IL). Samples (30 μg) were separated on a 12% acrylamide gel and transferred to a nitrocellulose membrane (Hybond; GE Healthcare, Piscataway, NJ). The membrane was probed with antibodies to Gnat2 (rabbit antibovine G_αt2, 1:3000; Santa Cruz Biotechnology, Inc.) and Gapdh (1:3000, mouse antirabbit; Sigma), incubated with fluorescent secondary antibodies (goat antirabbit 680, goat antimouse 800; Li-Cor Biosciences, Lincoln, NE), and analyzed (Odyssey infrared imager, Li-Cor Biosciences). 

Data Analysis

ERG data were modeled and analyzed using Excel (Microsoft Office Excel 2007; Microsoft, Redmond, WA), while graphing and statistical analysis were performed by using GraphPad Prism 5 (GraphPad Software, San Diego, CA). Where appropriate, data are presented as mean ± SEM and were analyzed using either a one-way ANOVA or 2-way ANOVA incorporating a Bonferroni post hoc using an α of 0.05. 

Results

Identification of a Novel, Naturally Occurring Mutation in the Gnat2 Gene

Mouse models are of particular use in investigating mechanisms of retinal disease, due in part to the relative ease of genetic manipulation. However, as highlighted recently, ^34,35 the presence of underlying background mutations can produce disease-related phenotypes that are not due to the altered gene of interest. While validating a new strain to the laboratory (Cx3cr1^GFP/GFP/BABL/c), a dramatic cone photoreceptor phenotype was observed that was unlike the phenotype previously reported for the Cx3cr1^GFP/GFP animal. ³⁶ Preliminary investigations suggested that this phenotype may be due to alterations in cone-specific transducin (Gnat2). The full-length sequence for the Gnat2 gene was amplified, and the product sequenced. A single A to G point mutation was observed at nucleotide 518 in exon 5 (see Fig. 1A), which resulted in an aspartic acid to glycine amino acid change at position 173 (Fig. 1B). This is the first description of this Gnat2 mutation, although a separate missense mutation (Gnat2^cpfl3 ; exon 6 c.598G>A, Asp200Asn) has been previously reported. ¹⁶  

Figure 1

 

Identification of the Gnat2^c.518A>G mutation. Genomic DNA was isolated from BALB/c, Gnat2^HET, and Gnat2^c.518A>G animals, and the Gnat2 coding region was amplified and sequenced. The genotyping results for WT (186 bp), Gnat2^HET (186, 105, and 81 bp), and Gnat2^c.518A>G (105 and 81 bp) are shown alongside the respective sequence chromatographs of a 13-bp region in exon 5 of Gnat2 showing the single A to G base pair substitution (A). The 200- and 100-bp standards are highlighted for size comparison. (B) The full-length exon 5 gene sequence for Gnat2 was sequenced for WT (BALB/c) and Gnat2^c.518A>G animals, and the corresponding translated sequence shows the Asp to Gly substitution (shaded area). The line above the sequence highlights the 13-bp region shown in (A).

Subsequent back crossing to the BALB/c resulted in the generation of the strain designated Gnat2^c.518A>G that was free of the Cx3cr1^GFP/GFP transgene. Since the mutation resulted in the generation of a novel Cac8I site within exon 5, PCR amplification followed by restriction endonuclease digest allowed for genotyping (see Fig. 1A; WT, 186 bp; Gnat2^HET, 186, 105, and 81 bp; Gnat2^c.518A>G, 105 and 81 bp). 

Retinal Function in the Gnat2^c.518A>G Mouse

In order to assess the effect of this missense mutation on retinal function, ERGs were performed on BALB/c and Gnat2^c.518A>G animals at 3, 6, 9, and 12 months of age. Representative ERG waveforms from 3-month-old animals are shown for the cone (Fig. 2A) and rod (Fig. 2B) responses. While the rod response appears normal between the two strains, there was no appreciable cone pathway response in the Gnat2^c.518A>G animals. This lack of a cone response was evident at all time points investigated, with cone b-wave amplitudes significantly reduced compared to WT animals (Fig. 2C, 2-way ANOVA, P < 0.0001). The BALB/c animals showed an age-related decline in the cone b-wave amplitude over time (Fig. 2C, 1-way ANOVA, P < 0.01), which has been previously reported. ³⁷ Unlike the cone response, the rod response was similar between the two strains at 3, 6, and 9 months. However, at 12 months of age the rod a-wave amplitude was decreased by 14% (Fig. 2D, 518 ± 11 μV vs. 446 ± 18 μV, 2-way ANOVA, P < 0.05). There was no difference in the rod b-wave response at any time point investigated (Fig. 2E). Both rod a- and b-wave responses showed age-related decreases in amplitude for BALB/c (1-way ANOVA, P < 0.0001 and P < 0.001, respectively) and Gnat2^c.518A>G animals (1-way ANOVA, P < 0.0001). Such age-related declines in rod pathway function have been previously reported. ^37,38 All other ERG parameters (sensitivity, implicit time, OPs) were unchanged in the Gnat2^c.518A>G animal. Overall, these data indicate that cone pathway function is absent in the Gnat2^c.518A>G mice, while rod photoreceptor function only becomes affected by 12 months of age. 

Figure 2

 

Retinal function in the Gnat2^c.518A>G mouse. The ERG was used to assess retinal function in WT (BALB/c) and Gnat2^c.518A>G animals. Representative ERG traces for 3-month-old WT and Gnat2^c.518A>G animals are shown for the cone (A) and rod (B) responses. The ERG responses were analyzed at 3, 6, 9, and 12 months of age to investigate the cone b-wave (PII, [C]) and rod a- and b-wave (PIII and PII, [D] and [E], respectively). Data are presented as mean ± SEM, n > 9. *P < 0.05, ****P < 0.0001.

Cone Transducin Expression in the Gnat2^c.518A>G Mouse

Sections from BALB/c and Gnat^c.518A>G retinae were labeled with an antibody to cone transducin to determine whether the presence of the mutation resulted in altered retinal expression (rabbit anti-G_αt2, 1:10,000; Santa Cruz Biotech). As can be observed in Figures 3A and 3C, control animals showed transducin labeling in photoreceptor outer segments (OS). Double labeling with the cone marker PNA showed complete colocalization (data not shown). Unlike the control sections, no transducin labeling was present in the Gnat2^c.518A>G animals at either 3 or 12 months of age (Figs. 3B, 3D, respectively). Higher concentrations of the transducin antibody (1:100 and 1:1000) were also tested on Gnat2^c.518A>G retinal sections; however, no specific labeling could be detected (data not shown). In order to probe whether this lack of Gnat2 expression in the tissue sections was due to an altered structural epitope, Western blots were performed on 3-month tissue samples using the same Gnat2 antibody. As can be observed in Figure 3E, two bands approximating the correct size (40 kDa) were observed in both the control and Gnat2^c.518A>G samples. The presence of two bands may reflect posttranslational modifications such as tyrosine phosphorylation, which is known to occur in the α subunit of transducin. ³⁹  

Figure 3

 

Expression of cone transducin in the Gnat2^c.518A>G mouse. Cone transducin expression (Gnat2, red) was assessed in retinal sections from BALB/c (A, C) and Gnat2^c.518A>G (B, D) animals at 3 (A, B) and 12 (C, D) months of age, with cell nuclei labeled with 4′,6-diamidino-2-phenylindole (DAPI; blue). Western blots (E) were performed on retinal extracts from 3-month-old BALB/c and Gnat2^c.518A>G animals, while Gapdh was used as a protein loading control. Gnat2 mRNA expression was quantified (F) using qPCR and was expressed relative to the housekeeping gene Hprt. Data are presented as mean ± SEM, n = 9, *P < 0.05. Scale bar: 10 μm.

The presence of Gnat2 protein in the Gnat2^c.518A>G retinae was supported by the qPCR data (Fig. 3F), which showed no alteration in Gnat2 copy number at 3 months, while retinae from older Gnat2^c.518A>G animals exhibited a 27% reduction in expression (12 months; 7.0 ± 0.5 copies Gnat2/copy Hprt versus 5.1 ± 0.6 copies Gnat2/copy Hprt). Other key genes involved in the phototransduction pathway were also quantified, with the expression levels of CngA3 and CngB3, both of which have been implicated in achromatopsia, ^15,17 not altered in the Gnat2^c.518A>G retina (data not shown). Thus, cone transducin immunolabeling is absent in the Gnat2^c.518A>G retinae, while Gnat2 gene expression is only decreased by 12 months of age. 

Opsin Localization and Expression in the Gnat2^c.518A>G Mouse

Previous work on the CngA3^−/− model of achromatopsia indicated altered opsin expression and mislocalization. ²² In order to determine whether cone opsin localization and/or expression was altered in the Gnat2^c.518A>G retina, sections were labeled with antibodies to M- and S-opsin. As observed in Figure 4A, M-opsin expression was confined to the cone OS in the BALB/c control animals at 3 months, while the Gnat2^c.518A>G retina showed opsin mislocalized to the photoreceptor cell bodies in the outer nuclear layer (ONL) and the terminals in the outer plexiform layer (OPL; Fig. 4B). By comparison, more extensive mislocalization was found in the 12-month Gnat2^c.518A>G tissue (Fig. 4D), while the age-matched control tissue showed only occasional instances of M-opsin expression in the photoreceptor cell bodies (see arrow, Fig. 4C). The expression of the corresponding gene (Opn1mw) was not altered at either 3 or 12 months (Fig. 4E). Unlike M-opsin, neither the localization of S-opsin, nor the expression of its gene (Opn1sw) was altered (data not shown). Rhodopsin also showed similar localization and gene expression between BALB/c and Gnat2^c.518A>G retinae (data not shown). Therefore, despite normal opsin gene expression profiles, Gnat2^c.518A>G retinae exhibit early M-opsin mislocalization, which increases in severity with age. 

Figure 4

 

Localization and expression of cone opsin in the Gnat2^c.518A>G mouse. M-opsin localization (green) was assessed in retinal sections from BALB/c (A, C) and Gnat2^c.518A>G (B, D) animals at 3 (A, B) and 12 (C, D) months of age. The mRNA expression of Opn1mw was quantified (E) using qPCR and was expressed relative to the housekeeping gene Hprt. Data are presented as mean ± SEM, n = 9. Scale bar: 10 μm.

Cone Photoreceptor Loss in the Gnat2^c.518A>G Mouse

In order to determine whether the loss of cone function resulted from cone photoreceptor degeneration in the Gnat2^c.518A>G retina, cone photoreceptors were labeled with PNA (green, Figs. 5A–D). At 3 months of age there appeared to be little qualitative change in cone number between the BALB/c and Gnat2^c.518A>G retinae (Figs. 5A, 5B), yet at 12 months there were fewer labeled cone photoreceptors in the Gnat2^c.518A>G eyes. When cone photoreceptor number was quantified, there was no alteration at 3 months of age, while at the 12-month time point there was a 27% decrease in the number of cone photoreceptors (88 ± 2 vs. 64 ± 5 cones/mm retina, 2-way ANOVA, P < 0.001). Estimation of retinal ONL thickness and rhodopsin labeling showed no change at either 3 or 12 months (data not shown). 

Figure 5

 

Cone photoreceptor loss in the Gnat2^c.518A>G mouse. Cone photoreceptors were labeled in WT (BALB/c, [A, C]) and Gnat2^c.518A>G (B, D) retinae with PNA (green) at 3 (A, B) and 12 (C, D) months of age. The number of cones was subsequently quantified (E) per millimeter of retina. Data are presented as mean ± SEM, n = 6, ***P < 0.001. Scale bar: 10 μm.

Retinal Remodeling and Gliosis in the Gnat2^c.518A>G Mouse

Loss of functional photoreceptors is known to result in retinal remodeling of synaspses. ^40,41 Specific immunocytochemical cellular markers were used to assess remodeling of downstream retinal neurons, and retinal remodeling was assessed at 3 and 12 months to determine whether such reorganization occurred prior to or coincident with cone photoreceptor loss (see Fig. 5E). 

Horizontal cells were labeled with the marker calbindin (green), ⁴² while an antibody to vesicular glutamate transporter 1 (VGlut1; red) was used to label photoreceptor terminals in the OPL (Fig. 6). ⁴³ In the 3-month BALB/c tissue, horizontal cells and photoreceptor terminals were confined to the OPL, while in the corresponding Gnat2^c518A>G retina, photoreceptor terminals were occasionally displaced into the ONL with horizontal cell processes also extending into the ONL and appearing to make contact (Figs. 6A, 6B, respectively). In the 12-month tissue, there was increased photoreceptor terminal and horizontal cell process displacement into the ONL (Fig. 6D), while the age-matched control only showed occasional remodeling (BALB/c, see arrow, Fig. 6C). 

Figure 6

 

Horizontal cell remodeling in the Gnat2^c.518A>G mouse. Horizontal cell remodeling was assessed in WT (BALB/c) and Gnat2^c.518A>G retinae by using immunohistochemistry. Retinal sections from 3-month-old (A, B) and 12-month-old (C, D) animals were stained with antibodies to calbindin (green) and vesicular glutamate transporter 1 (VGlut, red) to label horizontal cells and photoreceptor terminals, respectively. Gnat2^c.518A>G (B, D) animals showed displaced horizontal cell processes and photoreceptor terminals compared to age-matched controls (BALB/c, [A, C]). INL, inner nuclear layer. Scale bar: 10 μm.

Cone bipolar cells were labeled with ZNP, a marker for Type 2 OFF- and Type 6 ON-CBCs (green), ^44,45 while photoreceptor terminals were again labeled with VGlut1 (red). Control animals (Figs. 7A, 7C) showed few alterations in cone bipolar cell morphology and photoreceptor terminal interaction at 3 and 12 months of age. In contrast, 3-month-old Gnat2^c.518A>G retinae exhibited displaced photoreceptor terminals, with cone bipolar cell dendrites extending into the ONL (Fig. 7B), similar to the horizontal cell remodeling. This became more pronounced at 12 months of age (Fig. 7D). While the ERG data showed no alteration in the rod pathway of Gnat2^c.518A>G animals at 3 months (Figs. 2D, 2E), rod bipolar cells labeled with protein kinase C alpha (PKCα, green) showed occasional evidence of dendrite extension into the ONL in close approximation to the displaced photoreceptor terminals (red, Fig. 8B) when compared to the control (Fig. 8A). At 12 months when a rod ERG deficit was evident, there was more extensive dendrite extension into the ONL (Fig. 8D; age-matched control, Fig. 8C). 

Figure 7

 

Cone bipolar cell remodeling in the Gnat2^c.518A>G mouse. Cone bipolar cell remodeling was assessed in WT (BALB/c) and Gnat2^c.518A>G retinae using immunohistochemistry. Retinal sections from 3-month-old (A, B) and 12-month-old (C, D) animals were stained with antibodies to ZNP (green) and vesicular glutamate transporter 1 (VGlut, red) to label cone bipolar cells and photoreceptor terminals, respectively. Gnat2^c.518A>G (B, D) animals show aberrant cone bipolar cell–photoreceptor associations in the ONL compared to age-matched controls (BALB/c, [A, C]). Scale bar: 10 μm.

Figure 8

 

Rod bipolar cell remodeling in the Gnat2^c.518A>G mouse. Rod bipolar cell remodeling was assessed in WT (BALB/c) and Gnat2^c.518A>G retinae using immunohistochemistry. Retinal sections from 3-month-old (A, B) and 12-month-old (C, D) animals were stained with antibodies to protein kinase Cα (PKCα, green) and vesicular glutamate transporter 1 (VGlut, red) to label rod bipolar cells and photoreceptor terminals, respectively. Gnat2^c.518A>G (B, D) animals showed the presence of rod bipolar cell–photoreceptor associations in the ONL compared to age-matched controls (BALB/c, [A, C]). Scale bar: 10 μm.

Müller cell gliosis is commonly used as a measure of retinal stress and is generally increased during retinal degeneration. Müller cells within the BALB/c and Gnat2^c.518A>G retinae were labeled with glutamine synthetase (green, Figs. 9A–D), while GFAP (red) was used as a measure of gliosis. In the BALB/c and Gnat2^c.518A>G 3-month tissue (Figs. 9A, 9B), GFAP labeling was restricted to retinal astrocytes, with no evidence of Müller cell involvement. At 12 months of age, there was extensive Müller cell gliosis within the Gnat2^c.518A>G retinae compared to the age-matched control (Figs. 9C, 9D). 

Figure 9

 

Müller cell gliosis in the Gnat2^c.518A>G mouse. Retinal gliosis was assessed in 3 (A, B) and 12 (C, D) month old WT (BALB/c, [A, C]) and Gnat2^c.518A>G (B, D) retinae using immunohistochemistry. Sections were stained with antibodies to glutamine synthetase (green) and GFAP (red). While 3 month old WT and Gnat2^c.518A>G animals show similar retinal staining (A, B), 12 month old Gnat2^c.518A>G animals exhibit greater Müller cell gliosis than that observed in the respective age-matched control ([C] compared to [D]). GCL, ganglion cell layer. Scale bar: 10 μm.

Discussion

This study detailed a naturally occurring mutation in the cone transducin gene (Gnat2) and characterized the retinal phenotype of mice carrying this alteration. Sequencing of the full-length Gnat2 gene identified an A to G point mutation at nucleotide 518 that resulted in an aspartic acid to glycine substitution. Mice harboring this mutation had no cone response, showed evidence of altered transducin protein structure and M-opsin mislocalization, and developed cone degeneration by 12 months of age. Prior to cone loss, Gnat2^c.518A>G retinae showed displaced photoreceptor terminals and horizontal and cone bipolar cell remodeling, which increased in severity in older animals. There was also evidence of rod pathway involvement, with rod bipolar cell remodeling occasionally observed at 3 months of age and a decrease in rod photoreceptor function at 12 months of age. 

A spontaneous, natural mutation was observed in the background of the Cx3cr1^GFP/GFP/BALB/c transgenic mouse line. Automated sequencing of the full Gnat2 coding sequence identified a single A to G point mutation at nucleotide 518 in exon 5, which resulted in an amino acid substitution at codon 173 (D173G). This is the first report of this missense mutation in the literature and the second report of a Gnat2 mouse model of achromatopsia. Previously, Chang et al. ¹⁶ characterized a mutation in exon 6 of Gnat2 that resulted in a D200N substitution (Gnat2^cpfl3 ). While the c.518A>G mutation described in this report has a retinal phenotype similar to that found in individuals suffering achromatopsia, this particular missense mutation has not been identified in human patients, with the Gnat2 mutations identified thus far arising from insertions/deletions and truncations. ^9,10,12,13  

Subsequent backcrossing to BALB/c generated the Gnat2^c.518A>G strain, which exhibited no measurable cone response at 3 or 12 months of age. A lack of cone response is one of the key clinical indicators for individuals suffering complete achromatopsia. ⁴⁶ A similar loss of cone function has been reported in other animal models of achromatopsia, with the Gnat2^cpfl3 , CNGA3^−/−, and cpfl1 models all showing complete loss of cone function, ^11,15,16 while the CNGB3^−/− model only showed a 75% reduction in photopic ERG at 1 month. ¹⁷ In addition to the loss of cone function, the Gnat2^c.518A>G retinae also showed a reduced rod photoreceptor (a-wave) response at 12 months. The literature on rod pathway involvement in achromatopsia is less clear. Reflecting the current data, the CNGA3^−/− mouse model showed a rod a-wave decrease from 9 months (−20%), in addition to an early rod b-wave decrease (−15% at 1 month), while Chang et al. ¹⁶ alluded to reduced rod-mediated responses with age in the Gnat2^cpfl3 mouse, despite reporting near normal responses at 9 months. ²³ On the other hand, other achromatopsia models (cpfl1, CNGB3^−/−) exhibit no alteration in scotopic ERG responses. ^11,17 The variation in the animal model data is somewhat reflected in the human disease, for while patients suffering from achromatopsia generally have normal rod function, there are reports of functional decreases in the rod pathway. ^5,47  

While Gnat2^c.518A>G cone photoreceptors essentially exhibited a Gnat2^−/− phenotype in retinal section, Gnat2 protein was still present via Western blot. Such a result suggests that the D173G substitution altered the three-dimensional structure of Gnat2 to such an extent that the antibody no longer recognized native protein. This possibility is supported by the qPCR data, which showed normal transducin gene expression at 3 months, with Gnat2 gene expression only decreased at 12 months when cone loss had occurred. Thus, these data suggest that that the lack of cone function in the Gnat2^c.518A>G is due to altered transducin protein structure. However, in order to fully validate this, a high resolution three-dimensional structure is required. Unfortunately, the protein structure of Gnat2 has yet to be resolved; however, the structure of rod transducin (Gnat1) has been determined down to 1.7Å. ⁴⁸ Since Gnat1 and Gnat2 share a high sequence identity ⁴⁹ and are functionally interchangeable, ⁵⁰ it is likely that they share a similar structure. Thus, based on the published Gnat1 structure, the D173G mutation identified in this study would impact on an area bordering the GDP binding site (α helix 9, 1TAD, RCSB Protein Data Bank). Since the binding of GDP within the α subunit of transducin and subsequent opsin-mediated GTP exchange are critical for photoreceptor hyperpolarization, such a structural alteration to this binding site would explain the lack of cone function in the Gnat2^c.518A>G. 

In addition to the altered transducin structure, Gnat2^c.518A>G animals also showed M-opsin mislocalization at both 3 and 12 months of age, with protein expression localized around the somata and in the cone photoreceptor terminals. This altered expression pattern was not observed for S-opsin; furthermore, opsin gene expression (Opn1mw, Opn1sw, Rho) was not changed. While similar M-opsin mislocalization has been reported for the CNGA3^−/− and CNGB3^−/− achromatopsia models, these studies also show S-opsin mislocalization. ^18,22 The reason for the specific M-opsin effect in the current data is unclear as both types of cone photoreceptor express Gnat2. 

Opsin mislocalization is known to precede/accompany photoreceptor degeneration, with the mislocalized protein itself playing an important role in photoreceptor cell death, possibly via increasing endoplasmic reticulum stress. ^51,52 While cone numbers were normal at 3 months in the Gnat2^c.518A>G animals, cone degeneration and extensive Müller cell gliosis were observed at 12 months. This later degeneration correlates with the data from the Gnat2^cpfl3 mouse, which shows normal cone numbers at 14 weeks. ¹² The cone degeneration in this model is, however, slower and less aggressive than that reported for the CNGA3^−/− and CNGB3^−/− models, both of which show almost a 50% loss by 1 month and early Müller cell gliosis. ^18,22 While significant rod photoreceptor cell loss has been reported in the CNGA3^−/− retinae (15% loss of ONL), ²³ there was no evidence of such widespread degeneration in the Gnat2^c.518A>G animal. 

Preceding and coincident with the cone photoreceptor cell loss, horizontal and cone bipolar cell (type 2 OFF, type 6 ON) remodeling was also observed. Horizontal cell neurites and cone bipolar cell dendrites were observed to occasionally extend into the ONL by 3 months, with the extent of remodeling dramatically increased at 12 months. Retinal remodeling is commonly observed after degeneration and similar horizontal and bipolar cell extension has been detailed in the CNGA3^−/− model ^21,22 ; however, cone bipolar cell dendrite extension was restricted to the outer OPL, rather than the ONL. Within the Gnat2^c.518A>G retinae, the remodeled horizontal and bipolar cells were seen to be in close approximation with displaced photoreceptor terminals. While VGlut1 staining doesn't distinguish between cone and rod photoreceptors, no displaced terminals were observed using the cone-specific PNA label, suggesting these terminals were displaced rod photoreceptors. Such an increased proportion of displaced rod photoreceptors at 12 months may help explain the reduced rod function observed at this time point. Supporting this, previous work in models lacking cone function, including the CNGA3^−/−, has shown that neurons within the cone pathway remodel to receive input from rod photoreceptors. ^21,53 In addition to the cone-related remodeling, rod bipolar cells also showed extension of dendrites into the ONL as early as 3 months. While this appears to be the first evidence of rod bipolar cell remodeling in a model of achromatopsia, changes in the rod pathway have been implicated in other achromatopsia models, ^16,23 as mentioned previously, and other studies have shown similar rod bipolar cell remodeling in retina lacking cone function. ⁵³  

In summary, the Gnat2^c.518A>G mouse contains a naturally occurring missense mutation in exon 5 of the cone transducin gene that results in a D173G substitution. This mutation leads to a complete lack of a functional cone response in these mice, most likely due to an altered cone transducin structure. The retinal phenotype is characterized by mislocalized M-opsin, retinal remodeling, and late cone degeneration. This cone-specific mutation also leads to rod pathway remodeling and altered rod pathway function in older animals. The Gnat2^c.518A>G shows all the hallmark signs of a model of achromatopsia. 

Acknowledgments

The authors thank Paul McMenamin for providing the original Cx3cr1^gfp/gfp/BALB/c strain. 

Supported by the National Health and Medical Research Council (NHMRC) of Australia Grant 566814, the American Health Assistance Foundation (AHAF) Grant M2010010, and the Victorian State Government (Victoria's Science Agenda grant). 

Disclosure: A.I. Jobling, None; K.A. Vessey, None; M. Waugh, None; S.A. Mills, None; E.L. Fletcher, None 

Footnotes

 AIJ and KAV are joint first authors.