Abstract
Transducin is a heterotrimeric G-protein that is a component of the phototransduction cascade in rod and cone photoreceptor cells of the retina. Gnat1, a rod-specific transducin α-subunit, regulates light/dark adaptation by changing its subcellular localization depending on light. Our previous study revealed that Gnat1 translocation in rod photoreceptor cells under light/dark conditions requires E3 ligase Klhl18-mediated ubiquitination and degradation of Unc119, a Gnat1-binding protein. A mutation in the human UNC119 gene is associated with cone–rod dystrophy (CRD); however, the underlying pathological mechanism remains unclear. In this study, we generated and analyzed Unc119-deficient (Unc119−/−) mice. We found that the retinas of Unc119−/− mice of both sexes exhibited progressive photoreceptor degeneration, resembling CRD in humans. We also found that Unc119 interacts with Gnat2 in cone photoreceptor cells and that Unc119 is essential for the translocation of Gnat2 to the outer segment in cone photoreceptor cells. RNA-seq and subsequent bioinformatics analysis revealed the predicted activation of the JAK-STAT and NF-κB pathways in the Unc119−/− retina. Treatment of Unc119−/− mice with curcumin, an inhibitor of the JAK-STAT and NF-κB pathways, suppressed inflammation and cone photoreceptor cell degeneration in Unc119−/− retinas. Furthermore, a human CRD-associated UNC119 mutant protein competitively inhibited the interaction between UNC119 and GNAT1 or GNAT2. Taken together, the current study suggests that UNC119 dysfunction leads to CRD by affecting the JAK-STAT and NF-κB pathways and may advance our understanding of the pathological mechanisms of CRD.
Significance Statement
Unc119 functions in light/dark adaptation by modulating the subcellular localization of transducin, a phototransduction G-protein, in rod photoreceptor cells. A human UNC119 gene mutation is associated with cone–rod dystrophy (CRD), manifesting as cone degeneration followed by rod degeneration; however, the underlying pathological mechanism remains unclear. In this study, we found that Unc119-deficienct mouse retina exhibited progressive photoreceptor cell degeneration, resembling CRD. JAK-STAT and NF-κB inflammatory pathways were activated in the Unc119−/− retina, and treatment of Unc119−/− mice with curcumin, an inhibitor of JAK-STAT and NF-κB pathways, suppressed cone photoreceptor degradation. Furthermore, a human CRD-associated UNC119 mutant protein competitively inhibited the interaction between UNC119 and a transducin component. This study advances our understanding of the pathological mechanisms that underlie CRD.
Introduction
Unc119 was initially identified as a causative gene for Caenorhabditis elegans mutants associated with reduced motility, feeding behavior, and chemosensation in this organism (Maduro and Pilgrim, 1995). Unc119 is evolutionarily conserved across species, from C. elegans to mice and humans, whose homologs are referred to as Mrg4 and HRG4, respectively (Higashide et al., 1996; Ishiba et al., 2007). Unc119 is notably expressed in photoreceptor cells of the retina (Higashide et al., 1998; Swanson et al., 1998) and is also present in T lymphocytes, leukocytes (eosinophils), lung fibroblasts, adrenal glands, cerebellum, and kidney (Swanson et al., 1998; Gorska et al., 2004; Vepachedu et al., 2009). Unc119 interacts with Gnat1, a rod-specific transducin α-subunit, in rods (Zhang et al., 2011). N-terminal acylation of Gnat1 is essential for Gnat1 subcellular translocation in rods (Kerov et al., 2007) and for interaction with Unc119 (Zhang et al., 2011). Gnat1 relocates from the outer segment (OS) to the cell body in response to light stimulation and plays a critical role in light adaptation (Brann and Cohen, 1987; Philp et al., 1987; Whelan and McGinnis, 1988). Studies on Unc119-deficient mice have shown that Gnat1 localizes to both the OS and the cell body of rods, even in the dark, suggesting that Unc119 plays an essential role in regulating Gnat1 subcellular localization during light/dark adaptation (Zhang et al., 2011). In contrast, the subcellular localization of Gnat2, a cone-specific transducin α-subunit, is known to be stable in the OS under natural conditions of illumination (Lobanova et al., 2010). This indicates that cones have a different mechanism from rods in avoiding response saturation to bright light. However, the mechanism by which Gnat2 localizes to the OS has not yet been reported. Additionally, our previous study found that the light-induced subcellular localization of Gnat1 is modulated by the E3 ligase Klhl18, which ubiquitinates and degrades Unc119 (Chaya et al., 2019). Previous research showed that Unc119 localizes at the photoreceptor synapse and interacts with synaptic proteins. Unc119 facilitated glutamate release at rod synapses (Fehlhaber et al., 2023). On the other hand, the functional role of Unc119 in cones remains unclear.
Retinal neurons do not regenerate once they degenerate, leading to irreversible blindness (Gasparini et al., 2019). Retinitis pigmentosa is a representative retinal degeneration disease in which rod degeneration develops first, followed by the secondary degeneration of cones. In contrast, cone–rod dystrophy (CRD) is a progressive disorder characterized by initial cone degeneration followed by subsequent rod degeneration. CRD presents with symptoms, such as decreased visual acuity, reading difficulties, photophobia, and dyschromatopsia. As the disease progresses and rod function declines, affected individuals may experience reduced night vision and loss of peripheral visual fields (Hamel, 2007). Approximately one in 4,000 people are affected by this disease (Tsang and Sharma, 2018). At least 48 genes, including UNC119, have been reported to be associated with CRD (Birtel et al., 2018). Previous studies have identified several UNC119 mutations in patients with CRD and macular disease, including a heterozygous p.Lys57Ter (K57X) nonsense mutation (Kobayashi et al., 2000), a heterozygous p.Asp87Asn missense mutation (Huang et al., 2013), a heterozygous c.7delG (p.Val3*) frameshift mutation (de Castro-Miro et al., 2016), and a heterozygous p.Glu201Ter nonsense mutation (Zenteno et al., 2023). Transgenic mice overexpressing the K57X mutation exhibited a progressive decrease in b-wave amplitude in electroretinograms (ERGs) and thinning of the outer nuclear layer (ONL; Kobayashi et al., 2000). However, the mechanism of preferential degeneration in cones caused by CRD associated with UNC119 mutation remains unclear. In this study, we investigated the functional role of Unc119 in a mouse model and the pathological mechanisms of CRD induced by Unc119 mutations.
Materials and Methods
Animal care
All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the procedures were approved by the Institutional Safety Committee on Recombinant DNA Experiments (approval ID 04913) and Animal Experimental Committees of the Institute for Protein Research (approval ID R04-02-0), Osaka University, and performed in compliance with the institutional guidelines. Mice were housed in a temperature-controlled room at 22°C with a 12 h light/dark cycle. Freshwater and rodent diets were always available. All animal experiments were performed with mice of either sex at postnatal day 14 (P14), 1 month of age (1M), 6M, or 12M.
Generation of Unc119-deficient (Unc119−/−) and Nrl−/− mice
Unc119−/− and Nrl−/− mice were generated using the CRISPR/Cas9 system. Oligo DNAs for the gRNA sequence against mouse Unc119 and Nrl were cloned into the pX330 vector (Cong et al., 2013). The plasmid constructs were injected into C57/B6J fertilized eggs, which were then transferred into the uteri of pseudopregnant ICR female mice. Mutated individuals were selected using PCR and subjected to sequencing analysis. We obtained heterozygous mice with a total 65 base pair (bp) deletion in the Unc119 gene and heterozygous mice with a 239 bp deletion in the Nrl gene. Unc119−/− mice were backcrossed with Balb/c mice for more than six generations to generate Balb/c background Unc119−/− mice. The primer sequences for the gRNA expression plasmids and genotyping are listed in Table 1. The phenotypes of the Unc119−/− and Nrl−/− mice that we generated were comparable with those reported in the previous studies (Mears et al., 2001; Ishiba et al., 2007).
Primer sequence
Cell culture and transfection
HEK293T cells (RIKEN RCB, #RCB1637) were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) supplemented with penicillin (100 µg/ml) and streptomycin (100 µg/ml) at 37°C with 5% CO₂. Transfection was performed using the calcium phosphate method 24 h after seeding the HEK293T cells. Cell lysis was performed 48 h after transfection.
Plasmid construction
The cDNA fragments of mouse Unc119, Gnat1, Gnat2, Gnat1-G2A, and Gnat2-G2A and human UNC119, GNAT1, GNAT2, and UNC119-K57X were amplified by PCR using mouse or human retinal cDNA as a template and subcloned into pCAGGSII-3xFLAG, pCAGGSII-2xHA, and pGAGGSII-3xMyc vectors. Primer sequences used for amplification are shown in Table 1.
RT-PCR and quantitative RT-PCR (qRT-PCR) analyses
RT-PCR and qRT-PCR were performed as previously described (Tsutsumi et al., 2018). Total RNAs were isolated from mouse tissues, including the retina, cerebrum, cerebellum, brainstem, thymus, heart, lung, kidney, liver, spleen, muscle, intestine, ovary, and testis, using TRIzol RNA extraction reagent (Invitrogen). Total RNA (0.5 or 2 µg) was reverse transcribed into cDNA using the PrimeScript RT reagent or PrimeScript II reagent (Takara). The cDNAs were used as templates for RT-PCR reactions with rTaq polymerase (Takara). qRT-PCR analysis was performed using SYBR GreenER qPCR Super Mix Universal (Invitrogen) and Thermal Cycler Dice Real Time System Single MRQ TP700 (Takara) according to the manufacturer's instructions. Quantification was conducted using the Thermal Cycler Dice Real Time System software version 2.11 (Takara). The primer sequences used for qRT-PCR analysis are listed in Table 1.
Western blotting
Western blotting was performed as previously described (Kozuka et al., 2017). HEK293T cells were washed twice with PBS and lysed in lysis buffer supplemented with protease inhibitors (buffer A: 20 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, 1 mM PMSF, 2 µg/ml leupeptin, 5 µg/ml aprotinin, and 3 µg/ml pepstatin A). Mouse retinas were lysed in lysis buffer supplemented with protease inhibitors (buffer B: 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.5 mM EDTA, 1 mM PMSF, 2 µg/ml leupeptin, 5 µg/ml aprotinin, and 3 µg/ml pepstatin A). The samples were resolved by SDS-PAGE and transferred to PVDF membranes (Millipore) using a semidry transfer cell (Bio-Rad). Membranes were blocked with blocking buffer [3% skim milk and 0.05% Tween 20 in Tris-buffered saline (TBS)] for 1 h and incubated with primary antibodies overnight at 4°C. The membranes were washed three times with 0.05% Tween 20 in TBS for 10 min each and then incubated with secondary antibodies for 2–5 h at room temperature. The signals were detected using Chemi-Lumi One L (Nacalai) or Pierce Western Blotting Substrate Plus (Thermo Fisher Scientific). The following primary antibodies were used: mouse anti-FLAG M2 (1:5,000, Sigma, #F1804), rat anti-HA (1:5,000, Roche, #11-867-423-001, #60789700), rabbit anti-Myc-Tag (1:5,000, MBL, #562, #056), mouse anti-α-tubulin (1:5,000, Cell Signaling Technology, #DM1A, #T9026), rabbit anti-Unc119 (1:2,500, Chaya et al., 2019), rabbit anti-GNAT2 (1:2,500, Abcam, #ab97501), and mouse anti-GST (1:5,000, Nacalai, #04559-74). The following secondary antibodies were used: horseradish peroxidase-conjugated anti-mouse IgG (1:10,000, Zymed), anti-rat IgG (1:10,000, Jackson Laboratory), anti-rabbit IgG (1:10,000, Jackson Laboratory), and anti-guinea pig IgG (1:10,000, Jackson Laboratory).
Immunofluorescence analysis of retinal sections and whole-mount retinas
Immunofluorescence analysis of the retinal sections was performed as described previously (Yamamoto et al., 2020). The mouse eyes and eyecups were fixed in 4% paraformaldehyde (PFA) in PBS for 5 or 30 min at room temperature. The 30 min fixed samples were placed in 30% sucrose in PBS overnight at 4°C. The samples were embedded in Tissue-Tek O.C.T. Compound 4583 (Sakura). Frozen 20 µm sections placed on slides were dried overnight at room temperature. The tissue sections were washed three times with PBS. The samples were then incubated with blocking buffer (5% normal donkey serum and 0.1% Triton X-100 in PBS) for 1 h at room temperature and immunostained with primary antibodies in the blocking buffer overnight at 4°C. After washing with PBS, the samples were incubated with fluorescent dye-conjugated secondary antibodies and DAPI (1:1,000, Nacalai Tesque) for nuclear staining. The samples were washed three times with PBS and then coverslipped with gelvatol. For whole-mount immunostaining of the retina, each retina was gently peeled off from the sclera, rinsed in PBS, and fixed with 4% PFA in PBS for 30 min. After washing with PBS, the retinas were washed permeabilized by incubation in 0.1% Triton X-100 in PBS (PBST). The retinas were then immunostained with primary antibodies in PBST at 4°C for two overnight. Reactions with secondary antibodies in PBST were performed overnight at 4°C. The samples were washed three times with PBST and once with PBS and then coverslipped with Fluoro-KEEPER Antifade Reagent, Non-Hardening Type with DAPI (Nacalai Tesque, # 12745-74).
The primary antibodies used in this study were as follows: mouse anti-Rhodopsin (1:1,000, Sigma, # O4886), mouse anti-Pax6 (1:500, DSHB), mouse anti-Rom1 (1:250, a kind gift from Dr. R. Molday, University of British Columbia), rabbit anti-M-opsin (1:500, Millipore, #AB5405), goat anti-S-opsin (1:500, Santa Cruz, #sc-14363 or 1:500, ROCKLAND, #600-101-MP7S), guinea pig anti-Unc119 (1:250; Chaya et al., 2019), rabbit anti-Tα (1:500, Santa Cruz, #sc-389), rabbit anti-Chx10 (1:200, Koike et al., 2007), mouse anti-Pax6 (1:500, DSHB), rabbit anti-Calbindin (1:200, Calbiochem, #PC253L), mouse anti-Ctbp2 (1:500, BD Biosciences, #612044), guinea pig anti-Rbpms (1:1,000, Millipore, #ABN1376), rabbit anti-IbaI (1:500, WAKO, #019–19741), mouse anti-S100β (1:200, Sigma, #S-2532), rabbit anti-Gnat2 (1:500, Abcam, #ab97501), rabbit anti-Arr3 [1:500 (retinal sections) or 1:1,000 (whole mount), Sigma-Aldrich, #AB15282], mouse anti-GFAP (1:500, Sigma, #G3893-.2ML), rabbit anti-C1q (1:500, Abcam, #AB182451, #1006334-19), mouse anti-Socs3 (1:250, Proteintech, #66797-1-IG), and rabbit anti-NF-κB p65 (1:500, Cell Signaling Technology, #8342). Cy3-conjugated (1:500, Jackson ImmunoResearch Laboratories) and Alexa Fluor 488-conjugated (1:500, Sigma-Aldrich) secondary antibodies were used. The specimens were observed under a laser confocal microscope (LSM700 or LSM900, Carl Zeiss) at temperature of 22°C ± 3°C and humidity of <65%.
ERG recordings
ERGs were recorded as previously described (Sugita et al., 2020). Mice were adapted to the dark overnight. Mice were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg) diluted in saline (Otsuka). The pupils were dilated using topical 0.5% tropicamide and 0.5% phenylephrine HCl. ERG responses were measured using the PuREC system with LED electrodes (Mayo Corporation). Scotopic ERGs were recorded at four stimulus intensities ranging from −4.0 to 1.0 log cd s/m2. The mice were light adapted for 10 min before photopic ERGs were recorded on a rod-suppressing white background of 1.3 log cd s/m2. Photopic ERGs were recorded at four stimulus intensities ranging from −0.5 to 1.0 log cd s/m2. Eight and four responses were averaged for the scotopic recordings (−4.0 and −3.0 log cd s/m2, respectively). Sixteen responses were averaged for photopic recordings.
In situ hybridization
In situ hybridization was performed as previously described (Watanabe et al., 2015). Mouse embryos and eye cups were fixed overnight with 4% PFA in PBS at 4°C. The tissues were equilibrated in 30% sucrose in PBS overnight at 4°C, embedded in Tissue-Tek O.C.T. compound 4583 (Sakura), and frozen. Digoxigenin-labeled antisense and sense riboprobes for mouse Unc119 were synthesized by in vitro transcription using 11-digoxigenin UTPs (Roche). Unc119 cDNA fragments for in situ hybridization probes were produced by RT-PCR using mouse retinal cDNA as a template.
Toluidine blue staining
Toluidine blue staining of retinal sections was performed as previously described (Chaya et al., 2022). The mouse eyes were fixed in 4% PFA in PBS for 5 min at room temperature. The samples were embedded in Tissue-Tek O.C.T. Compound 4583 (Sakura). Frozen 20 µm sections placed on slides were dried overnight at room temperature. Retinal sections were washed with PBS and stained with 0.1% toluidine blue in PBS. After washing with PBS, the slides were coverslipped and observed immediately under a microscope. The ONL and inner nuclear layer + inner plexiform layer + ganglion cell layer thicknesses were measured and quantified using the ImageJ software (National Institutes of Health).
RNA-seq
RNA-seq analysis was performed as previously described (Yoshimoto et al., 2023) with some modifications. Total RNAs from three Unc119+/+ and three Unc119−/− mouse retinas at 1M were isolated using TRIzol RNA extraction reagent (Invitrogen). Sequencing was performed on an Illumina NovaSeq 6000 platform in 101-base single-end mode. Raw reads were mapped to mouse reference genome sequences (mm10) using the TopHat ver. 2.0.13, in combination with Bowtie 2 ver. 2.3.5.1 and SAMtools ver. 1.11. The number of fragments per kilobase of exon per million mapped fragments (FPKMs) was calculated using Cufflinks ver. 2.2.1. Using cutoff (fold change >1.4, <−1.4; p < 0.05, unpaired t test), heatmap visualization was conducted using the web tool Heatmapper. Gene set enrichment analysis (GSEA) was performed using R package clusterProfiler (Wu et al., 2021). Upstream regulator analysis was performed using IPA (Qiagen). IPA was used to predict the activated or inhibited transcription factors based on the observed differential gene expression profiles. RNA-seq analysis datasets are available in the Gene Expression Omnibus (GEO) database of NCBI (accession number GSE282222).
Immunoprecipitation
Immunoprecipitation assays were performed as previously described (Tsutsumi et al., 2022). HEK293T cells were cotransfected with plasmids expressing FLAG-, HA-, and Myc-tagged proteins. After 48 h of transfection, the cells were lysed in lysis buffer supplemented with protease inhibitors [20 mM Tris (hydroxymethyl) aminomethane (Tris)-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40 (NP-40), 1 mM ethylenediaminetetra-acetate (EDTA), 1 mM phenylmethylsulfonyl fluoride (PMSF), 2 µg/ml leupeptin, 5 µg/ml aprotinin, and 3 µg/ml pepstatin A]. Cell lysates were incubated with an anti-FLAG M2 affinity gel (Sigma-Aldrich) overnight at 4°C. The beads were washed five times with wash buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% NP-40, and 1 mM EDTA) and three times with TBS and then eluted with elution buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mg/ml 1× FLAG peptide) for 1 h at 4°C. The immunoprecipitated samples were incubated with sodium dodecyl sulfate (SDS) sample buffer for 30 min at room temperature and analyzed by Western blotting.
Pull-down assay and LC-MS/MS analysis
cDNA fragments encoding full-length mouse Unc119 were amplified by PCR and subcloned into the pGEX-4T-3 plasmid (GE Healthcare). GST-fused protein was expressed in Escherichia coli strain BL21 (DE3) and incubated with glutathione Sepharose 4B (GE HealthCare) for 3 h at 4°C. Nrl+/+ and Nrl−/− mouse retinas were lysed in lysis buffer supplemented with protease inhibitors (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.5 mM EDTA, 1 mM PMSF, 2 µg/ml leupeptin, 5 µg/ml aprotinin, and 3 µg/ml pepstatin A). The GST-fused proteins bound to glutathione Sepharose 4B were incubated with the retinal lysates for 3 h at 4°C, followed by five washes with wash buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% NP-40, and 1 mM EDTA) and five additional washes with TBS. The samples were then incubated with 0.1 M glycine, pH 1.5, on ice for 1 h. Eluted samples were analyzed by Western blotting or treated with trypsin and lysis endopeptidase were subjected to Orbitrap Exploris 480 (Thermo Fisher Scientific) for LC-MS/MS. The resulting data were processed using Mascot Distiller v2.8 (Matrix Science).
Drug administration
Curcumin (Sigma, #C7727) was dissolved in dimethyl sulfoxide (DMSO) at 200 mg/ml and stored at −20°C. The administration of curcumin to mice was performed as described previously (Varner et al., 2024), with some modifications. Curcumin was dissolved in DMSO (200 mg/ml) and diluted with sunflower seed oil (Sigma-Aldrich). The solvent, with or without curcumin (100 mg/kg), was administered subcutaneously to the mice daily from P21 for 7 d.
Quantification and statistical analysis
Quantification of immunofluorescence images was performed by ImageJ and ZEN software (Carl Zeiss) using maximum projection images (1,024 × 1,024 pixels) captured by confocal microscopy with a 20× lens. Each data point indicates the average of the measurements of the four images from one mouse. Statistical analyses were performed using GraphPad Prism version 9 (GraphPad Software). Data are presented as the mean ± SD. Statistical analyses were performed using the unpaired t test, two-way ANOVA Šídák's multiple-comparisons test, or one-way ANOVA multiple-comparisons test, as indicated in the figure legends. Asterisks indicate statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.
Results
Unc119 is expressed in developing and developed photoreceptor cells
To elucidate the functional role of Unc119 in cones and the pathophysiology underlying CRD, we examined Unc119 expression in mouse tissues. We conducted RT-PCR analysis using various tissues from 4-week-old mice and observed the ubiquitous expression of Unc119 (Fig. 1A). Next, we investigated the expression pattern of Unc119 in the mouse retina using in situ hybridization. We observed Unc119 signals in the neuroblastic layer (NBL) and ONL, where photoreceptor cells are located, from embryonic day 17.5 (E17.5) to P14 (Fig. 1B). These observations suggest that Unc119 is expressed in retinal photoreceptor cells during and after development.
Generation of Unc119−/− mice. A, RT-PCR analysis of Unc119 transcripts in mouse tissues at 4 weeks. B, In situ hybridization analysis of Unc119 transcripts in developing E17.5, P3, P9, and P14 retinas. GCL, ganglion cell layer; NBL, neuroblastic layer; ONL, outer nuclear layer; INL, inner nuclear layer. C, DNA sequences of exon 1 in wild-type and Unc119 mutant mice. Seven, one, and fifty-seven base pair (bp) deletions (total 65 bp deletion) resulted in a transcriptional frameshift and premature stop codon. D, Schematic representation of the Unc119 mutant allele. E, PCR products of 195 and 260 bp were amplified from the wild-type and Unc119 mutant alleles, respectively. F, RT-PCR analysis of Unc119 transcripts in Unc119+/+ and Unc119−/− retinas. β-Actin was used as a loading control. G, Western blot analysis of Unc119 protein in Unc119+/+ and Unc119−/− mouse retinas. α-Tubluin was used as a loading control. H, Immunofluorescence analysis of Unc119 protein in Unc119+/+ and Unc119−/− mouse retinas. I, J, Subcellular localization of Gnat1 in photoreceptor cells of Unc119+/+ and Unc119−/− retinas under dark- and light-adapted conditions at 1M. The measured Gnat1 signals in the ONL of the photoreceptors are shown. Data are presented as the mean ± SD. n = 3 mice. ***p < 0.001, n.s., not significant (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; OS, outer segment.
To investigate the functional role of Unc119 in the retina, we generated Unc119−/− mice by CRISPR-Cas9 genome editing (Fig. 1C–E). To confirm the loss of Unc119 expression in the retinas of Unc119−/− mice, we performed RT-PCR, Western blotting, and immunofluorescence analyses. These analyses showed a complete absence of Unc119 mRNA and protein expression in the retinas of the Unc119−/− mice (Fig. 1F–H). Previous studies have reported that Unc119 deficiency leads to the mislocalization of Gnat1 to the OS and ONL under dark-adapted conditions (Zhang et al., 2011). Unc119−/− mice exhibited Gnat1 mislocalization under dark-adapted conditions at 1M (Fig. 1I,J).
Unc119 deficiency decreases light responses in cones
To examine the electrophysiological properties of Unc119−/− retinas, we measured ERGs under dark-adapted (scotopic) and light-adapted (photopic) conditions at 1M. Under scotopic conditions, the a-wave amplitude reflects the responses of rods, whereas the b-wave amplitude reflects the combined activity of rod and rod bipolar cells. Under photopic conditions, the a-wave amplitude represents cone photoreceptor cell responses, and the b-wave amplitude reflects the combined activity of cone photoreceptors and cone bipolar cells. The previous study performed scotopic ERGs (Ishiba et al., 2007). We performed both scotopic and photopic ERGs in Unc119−/− mice. No significant differences were observed between Unc119+/+ and Unc119−/− mice in scotopic a- and b-waves and photopic a-wave amplitudes (Fig. 2A–E). However, a significant decrease in photopic b-wave amplitude was observed in Unc119−/− mice (Fig. 2D,F). Given that Unc119 is localized at photoreceptor synapses (Higashide et al., 1998; Fehlhaber et al., 2023), we tested signal transduction from photoreceptor cells to bipolar cells by measuring the b-wave implicit time in Unc119+/+ and Unc119−/− mice. We observed that the b-wave implicit time did not change significantly (Fig. 2G,H). These results suggest that Unc119 deficiency leads to a decline in cone photoreceptor function but does not significantly affect rod photoreceptor function at 1M and does not affect synaptic transmission from photoreceptors to bipolar cells.
Decreased cone light responses in Unc119−/− mice at 1M. A–F, ERGs analysis of Unc119−/− mice. ERGs were recorded from Unc119+/+ and Unc119−/− mice at 1M (n = 5 per each genotype). A, Representative scotopic ERGs elicited by four stimulus intensities (−4.0 to 1.0 log cd s/m2) from Unc119+/+ and Unc119−/− mice at 1M. B, C, Scotopic amplitudes of a- (B) and b-waves (C) are shown as a function of stimulus intensity. Data are presented as the mean ± SD. n = 5 per each genotype. n.s., not significant (two-way repeated-measures ANOVA, multiple comparisons). D, Representative photopic ERGs elicited by four stimulus intensities (−0.5 to 1.0 log cd s/m2) from Unc119+/+ and Unc119−/− mice at 1M. E, F, Photopic amplitudes of a- (E) and b-waves (F) are shown as a function of the stimulus intensity. Data are presented as the mean ± SD. n = 5 per each genotype. *p < 0.05, ***p < 0.001, n.s., not significant (two-way repeated-measures ANOVA, multiple comparison). G, H, Scotopic and photopic b-wave implicit times of ERGs were recorded from Unc119+/+ and Unc119−/− mice at 1M. The scotopic implicit times of the b-wave (G) reflect synaptic transmission from rod photoreceptors to rod bipolar rod cells. The photopic implicit times of the b-wave (H) reflect synaptic transmission from the cone photoreceptor to the cone ON bipolar cells. Data are presented as mean ± SD. n = 5 per each genotype. n.s., not significant (two-way repeated-measures ANOVA, multiple comparison). I, Toluidine blue staining of Unc119+/+ and Unc119−/− retinas at 1M. The thicknesses of the ONL, OPL, INL, IPL, and GCL were measured using ImageJ. Data are presented as the mean ± SD. n = 4 per each genotype. n.s., not significant (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; IPL, inner plexiform layer. J–R, Immunofluorescence analysis of retinal sections from Unc119+/+ and Unc119−/− at 1M using marker antibodies as follows: Rhodopsin (rod outer segments, J), M-opsin (M-cone outer segments, K), S-opsin (S-cone outer segments, L), Chx10 (bipolar cells, M), Pax6 (amacrine and ganglion cells, N), Calbindin (horizontal cells and a subset of amacrine cells, O), Ctbp2 (photoreceptor synapses, P), Rbpms (ganglion cells, Q), and S100β (Müller glial cells, R). The nuclei were stained with DAPI (blue). Rhodopsin length was measured in four images from the retina of one mouse. In one image, Rhodopsin length was measured at three points and averaged (J). The signal intensity of M-opsin and S-opsin in OPL were measured. Arrowheads indicate mislocalization of M-opsin and S-opsin signals (K, L). Data are presented as the mean ± SD. n = 3 per each genotype. n.s., not significant (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; OS, outer segment.
To investigate why the photopic b-wave amplitude in Unc119−/− mice at 1M decreased, we conducted a histological analysis of retinal sections from Unc119+/+ and Unc119−/− mice. Toluidine blue staining showed no significant differences in the retinal layer thickness between the two groups (Fig. 2I). Next, we performed immunofluorescence analysis using antibodies against Rhodopsin (a marker for rod outer segments), M-opsin (a marker for M-cone outer segments), and S-opsin (a marker for S-cone outer segments). Although the length of Rhodopsin and the number of M-opsin and S-opsin signals remained unchanged, mislocalization of M-opsin and S-opsin to the outer plexiform layer (OPL) was observed in the Unc119−/− retina (Fig. 2J–L). Mislocalization of the M-opsin and S-opsin signals to the OPL was not observed in the retina of Unc119−/− mice at P14 (Fig. S1). Immunofluorescence analysis using markers for other retinal cells, including Chx10 (bipolar cells), Pax6 (amacrine and ganglion cells), Calbindin (horizontal cells and some amacrine cells), Ctbp2 (photoreceptor synapses), Rbpms (ganglion cells), and S100β (Müller glial cells), showed no significant differences between the Unc119+/+ and Unc119−/− retinas (Fig. 2M–R, Fig. S2).
Unc119 deficiency causes progressive photoreceptor cell degeneration
Mutations in UNC119 have been identified in patients with CRD (Kobayashi et al., 2000; Huang et al., 2013; de Castro-Miro et al., 2016; Zenteno et al., 2023). To determine whether Unc119−/− mice exhibited a similar pathology, we conducted ERGs and histological analyses at 6M and 12M. While ERG analysis showed no significant differences in scotopic b-wave and photopic a-wave amplitudes between the Unc119+/+ and Unc119−/− mice at 6M, decreases in scotopic a-wave and photopic b-wave amplitudes were observed in the Unc119−/− mice (Fig. S3A–F). At 6M, the b-wave implicit time of ERGs in Unc119−/− mice showed no significant change compared with that in Unc119+/+ mice (Fig. S3G,H). In the histological analysis, toluidine blue staining revealed a reduction in ONL thickness in Unc119−/− mice at 6 M (Fig. S3I). Immunofluorescence analysis showed no significant change in Rhodopsin length but displayed a decrease in the number of M-opsin-, S-opsin-, and Arrestin 3 (Arr3)-positive cells, along with mislocalization of the M-opsin signals (Fig. S3J–M). A decrease in the number of cones was also observed in Unc119−/− mice by counting Arr3-positive OS signals using whole-mount tissues (Fig. S3N). These findings suggest that cones and a part of the rods are progressively degenerated in Unc119−/− mice at 6M.
ERG analysis at 12M revealed decreases in both scotopic a-wave and photopic b-wave amplitudes in Unc119−/− mice (Fig. 3A–F). The b-wave implicit time of ERGs in Unc119−/− mice was not significantly different from that in Unc119+/+ mice (Fig. 3G,H), suggesting that Unc119 deficiency leads to a decline in photoreceptor cell function but does not significantly affect synaptic transmission from photoreceptor cells to bipolar cells. Toluidine blue staining revealed further thinning of the ONL in Unc119−/− mice at 12M (Fig. 3I). Immunofluorescence analysis also demonstrated reductions in the thickness of Rhodopsin and in the number of M-opsin-, S-opsin-, and Arr3-positive cells (Fig. 3J–M). These findings suggest that Unc119 deficiency causes progressive photoreceptor cell degeneration, resembling CRD in humans.
Decrease in cone and rod light responses of Unc119−/− mice at 12M. A–F, ERG analysis of Unc119−/− mice at 12M. ERGs were recorded from Unc119+/+ and Unc119−/− mice at 12M (n = 5 and 4 mice, Unc119+/+ and Unc119−/−, respectively). A, Representative scotopic ERGs elicited by four stimulus intensities (−4.0 to 1.0 log cd s/m2) from Unc119+/+ and Unc119−/− mice at 12M. B, C, The scotopic amplitudes of a- (B) and b-waves (C) are shown as a function of the stimulus intensity. Data are presented as the mean ± SD. n = 5 and 4 mice (Unc119+/+ and Unc119−/− mice, respectively). **p < 0.01, ***p < 0.001, n.s., not significant (two-way repeated-measures ANOVA, multiple comparison). D, Representative photopic ERGs elicited by four stimulus intensities (−0.5 to 1.0 log cd s/m2) from Unc119+/+ and Unc119−/− mice at 12M. E, F, The photopic amplitudes of a- (E) and b-waves (F) are shown as a function of the stimulus intensity. Data are presented as mean ± SD. n = 5 and 4 mice (Unc119+/+ and Unc119−/−, respectively). *p < 0.05, **p < 0.01, n.s., not significant (two-way repeated-measures ANOVA, multiple comparison). G, H, Scotopic (G) and photopic (H) b-wave implicit times of ERGs were recorded from Unc119+/+ and Unc119−/− mice at 12M. Data are presented as the mean ± SD. n = 5 and n = 4 mice, Unc119+/+ and Unc119−/−, respectively. n.s., not significant (two-way repeated-measures ANOVA, multiple comparisons). I, Toluidine blue staining of Unc119+/+ and Unc119−/− retinas at 12M. The thicknesses of the ONL, OPL, INL, IPL, and GCL were measured using ImageJ. Data are presented as the mean ± SD. n = 3 and 4 (Unc119+/+ and Unc119−/−, respectively). *p < 0.05, ***p < 0.01, ****p < 0.001, n.s., not significant [unpaired t test (top bar graph) and one-way ANOVA, multiple comparisons (bottom bar graph)]. GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; IPL, inner plexiform layer; OS, outer segment. J–M, Immunofluorescence analysis of retinal sections from Unc119+/+ and Unc119−/− at 12M using marker antibodies as follows: Rhodopsin (rod outer segments, J), M-opsin (M-cone outer segments, K), S-opsin (S-cone outer segments, L), and Arr3 (cone photoreceptor cells, M). Nuclei were stained with DAPI (blue). Rhodopsin length was measured in four images from the retina of one mouse. In one image, Rhodopsin length was measured at three points and averaged (J). The number of M-opsin-, S-opsin-, and Arr3-positive cells was counted (K, L, M). Data are presented as mean ± SD. n = 3 and 4 (Unc119+/+ and Unc119−/−, respectively). *p < 0.05, **p < 0.01, ***p < 0.001 (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; OS, outer segment.
Unc119 interacts with Gnat2 in cones
To elucidate the functional mechanism of Unc119 in cones, we attempted to identify the interaction partners of the Unc119 protein in cones by performing a pull-down assay followed by LC-MS/MS analysis using Nrl−/− mouse retinas. Nrl is a transcription factor that regulates rod cell fate during development. Nrl deficiency causes loss of rod photoreceptor cell features and gain in cone photoreceptor cell characteristics (Swaroop et al., 1992; Mears et al., 2001). We generated Nrl−/− mice using the CRISPR/Cas9 system and performed Western blotting and immunofluorescence analysis of Nrl−/− retinas using antibodies against Rhodopsin, S-opsin, M-opsin, and Unc119. The loss of Rhodopsin signal and the increase in M-opsin and S-opsin signals suggest that the Nrl−/− mice also lost the characteristics of rods and gained those of cones. We observed that the localization and expression level of Unc119 were preserved in the Nrl−/− retina compared with those in the control retina (Fig. 4A, Fig. S4). We performed pull-down assay and LC-MS/MS analysis using three types of samples: GST protein reacted with Nrl−/− retinal lysates, GST-fused mouse Unc119 (mUnc119) reacted with lysis buffer, and GST-fused mUnc119 reacted with Nrl−/− retinal lysates. We used GST protein reacted with Nrl−/− retinal lysates and GST-fused mUnc119 reacted with lysis buffer as negative controls. Using normalized total spectra, 58 proteins were detected in the three types of samples; only four proteins, Gnat2, Ytdc2, Tba1a, and Hmmr, were detected specifically in GST-fused mUnc119 that reacted with Nrl−/− retinal lysates (Fig. 4B,C).
Unc119 interacts with Gnat2 in cones. A, Generation of Nrl−/− mice and Western blotting analysis of Nrl+/+ and Nrl−/− mouse retinas using the following antibodies: Rhodopsin, S-opsin, and M-opsin. α-Tubulin was used as a loading control. B, Scheme of pull-down assay and LC-MS/MS analysis. A pull-down assay was performed using GST and GST-fused Unc119 proteins. GST- and GST-fused Unc119 proteins were reacted with Nrl−/− retinal lysates or lysis buffer. Venn diagram of the number of identified proteins in the indicated experimental conditions by pull-down assay and subsequent LC-MS/MS analysis (left). C, Fifty-eight proteins were identified using total LC-MS/MS analysis. Four proteins, Gnat2, Ytdc2, Tba1a, and Hmmr, were detected only in GST-fused mUnc119 that reacted with Nrl−/− retinal lysates. The normalized total spectra of Gnat2 are shown. D, Immunoprecipitation of Unc119 with Gnat1 or Gnat2. Plasmids expressing FLAG-tagged Gnat1 or Gnat2, and HA-tagged Unc119 were cotransfected into HEK293T cells. Cell lysates were subjected to immunoprecipitation using an anti-FLAG antibody. The immunoprecipitated proteins were detected by Western blot analysis using anti-FLAG and anti-HA antibodies. E, F, Immunofluorescence analysis of Unc119+/+ and Unc119−/− retinas at 1M (D) and P14 (E), using an anti-Gnat2 antibody. Nuclei were stained with DAPI (blue). Gnat2 intensity was measured in the retinal sections of all photoreceptor layers (OS, IS, ONL, and OPL) using ImageJ software. Data are presented as the mean ± SD. n = 5 per each genotype at 1M and n = 3 per each genotype at P14. ***p < 0.001 (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; OS, outer segment. G, Western blotting analysis of Gnat2 protein in retinas from Unc119+/+ and Unc119−/− mice at P14 using an anti-Gnat2 antibody. α-Tubulin was used as a loading control. Relative Gnat2 protein levels in Unc119+/+ and Unc119−/− retinas were determined by quantification of Gnat2 band intensity (normalized to α-tubulin). Data are presented as the mean ± SD. n.s., not significant (unpaired t test); n = 5 mice per genotype. H, I, Immunoprecipitation analysis of Unc119 and Gnat1, Gnat1-G2A, Gnat2, or Gnat2-G2A. Plasmids expressing FLAG-tagged Gnat1, Gnat1-G2A, Gnat2, or Gnat2-G2A, and HA-tagged Unc119 were cotransfected into HEK293T cells. Cell lysates were subjected to immunoprecipitation with an anti-FLAG antibody. Immunoprecipitated proteins were detected by Western blot analysis with anti-FLAG (H) and anti-HA (I) antibodies.
Among these four proteins, based on expression pattern, we focused on the G-protein subunit alpha transducin 2 (Gnat2), a cone photoreceptor-specific transducin alpha subunit. The interaction between Unc119 and Gnat2 was examined by immunoprecipitation and pull-down assays (Fig. 4D, Fig. S5). In an immunoprecipitation assay, Unc119 interacted with Gnat2 as well as with Gnat1. This result suggests that Unc119 interacts with Gnat1 in rods and Gnat2 in cones. Immunofluorescence analysis with an antibody against Gnat2 was performed using Unc119+/+ and Unc119−/− mouse retinas at 1M and P14. Although the number of cone OS was not significantly different between the two groups at 1M and P14, the Gnat2 signal intensity in the OS, but not in the IS, ONL, and OPL, was reduced in the Unc119−/− retinas both at 1M and P14 (Fig. 4E,F; Fig. S6). To determine the expression level of Gnat2 protein, we performed Western blot analysis of retinas from Unc119+/+ and Unc119−/− mice at P14 using the anti-Gnat2 antibody. The Gnat2 protein expression level was not significantly different between Unc119+/+ and Unc119−/− retinas (Fig. 4G), suggesting that Unc119 is required for Gnat2 localization in the OS.
N-terminal glycine (G2) acylation of Gnat1 is essential for its interaction with Unc119 and for proper subcellular localization change of Gnat1 from the cell body to the OS (Kerov et al., 2007; Goc et al., 2008). To confirm whether N-terminal glycine (G2) acylation of Gnat2 is required for its interaction with Unc119, we generated a Gnat2-G2A mutant protein, in which G2 was replaced with alanine (G2A). We conducted immunoprecipitation analysis to test whether Unc119 interacts with Gnat1-WT, Gnat1-G2A, Gnat2-WT, or Gnat2-G2A (Fig. 4H,I). The immunoprecipitation results showed that Unc119 interacts with both Gnat1-WT and Gnat2-WT, but not with Gnat1-G2A or Gnat2-G2A, suggesting that G2 acylation of both Gnat1 and Gnat2 is essential for the interaction with Unc119.
NF-ĸB and JAK-STAT pathways are activated by Unc119 deficiency in the retina
To investigate the transcriptional profile alterations of Unc119−/− retinas, we conducted RNA-seq analysis of Unc119+/+ and Unc119−/− mouse retinas at 1M. We observed 80 downregulated genes and 286 upregulated genes in the Unc119−/− retina, except Unc119 (cutoff at average fold change >1.4 or less than −1.4; p < 0.05; Fig. 5A). We focused on four upregulated genes: glial fibrillary acidic protein (GFAP), complement component 1q subcomponent alpha polypeptide (C1qa), complement component 1q subcomponent beta polypeptide (C1qb), and complement component 1q subcomponent C chain (C1qc), since the upregulation of these genes is associated with glial activation and neurodegeneration (Sarthy et al., 1991; Jiao et al., 2018; Palko et al., 2022). To determine whether the expression of these genes was altered in Unc119-deficient retina in vivo, we immunostained retinal sections from Unc119+/+ and Unc119−/− mice with antibodies against GFAP and C1q. Increased GFAP and C1q signals were observed in the Unc119−/− retina, suggesting activation of Müller glial cells in the Unc119−/− retina (Fig. 5B,C). One of the downregulated genes was nucleoredoxin-like 2 (Nxnl2), also known as rod-derived cone viability factor 2 (Rdcvf2). Rdcvf2 is a thioredoxin-like protein secreted by rods that promotes cone survival (Chalmel et al., 2007). We examined the expression levels of Nxnl2 and nucleoredoxin-like 1 (Nxnl1), which encodes Rdcvf, in Unc119+/+ and Unc119−/− retinas using qRT-PCR analysis. We observed that the expression of Nxnl2 and Nxnl1 significantly decreased in Unc119−/− retinas compared with that in Unc119+/+ retinas (Fig. 5D). Next, we performed GSEA (Wu et al., 2021) to identify differentially expressed gene modules between Unc119+/+ and Unc119−/− mouse retinas (Fig. 5E). GSEA showed that genes related to immune responses were altered in the Unc119−/− mouse retina. Furthermore, we performed Ingenuity Pathway Analysis (IPA) to predict upstream transcription factors affecting gene expression in the Unc119−/− retina. IPA predicted that Irf1, Nfatc2, Ir3, Tp53, Nfkb1a, Rela, Irf7, Stat2, Stat3, and Klf6 were activated in the Unc119−/− retina (activation Z score >3.0; Fig. 5F). Almost all of the genes predicted to be activated are related to inflammation pathways, such as the JAK-STAT and NF-κB pathways. Notably, Irf1, a factor known to trigger the phosphorylation of STAT1 and thereby activate the JAK-STAT pathway (Guschin et al., 1995; Mishra and Ivashkiv, 2024), as well as Rela, a key member of the NF-κB protein family that governs NF-κB signaling (Ryseck et al., 1996), were activated. IPA networks showed that the transcription factors Irf1 and RelA are upstream regulators of multiple inflammatory genes (Fig. 5G,H). These results suggest that Unc119 deficiency directly or indirectly leads to activation of the JAK-STAT and NF-κB pathways, resulting in retinal inflammation.
Transcriptional changes by Unc119 deficiency in the retina. A, Heatmaps of differentially expressed genes (fold change >1.4, less than −1.4; p < 0.05, unpaired t test) between Unc119+/+ and Unc119−/− retinas. Normalized FPKM from the RNA-seq dataset was used for heatmap visualization. B, C, Immunofluorescence analysis of Unc119+/+ and Unc119−/− retinas at 1M using antibodies against GFAP and C1q. Nuclei were stained with DAPI (blue). The intensities of GFAP (B) and C1q (C) signals in Unc119+/+ and Unc119−/− retinas at 1M were measured. Data are presented as the mean ± SD. ∗∗p < 0.01 (unpaired t test). n = 3 mice per group. GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer. D, qRT-PCR analysis of Nxnl2 and Nxnl1 mRNA expression levels in the Unc119+/+ and Unc119−/− mouse retinas. Data are presented as the mean ± SD. ∗p < 0.05, ∗∗p < 0.01 (unpaired t test). n = 3 mice per group. E, Gene set enrichment analysis of upregulated genes in the Unc119−/− mouse retina. F, Ingenuity Pathway Analysis (IPA) to predict the upstream factors affecting gene expression changes (fold change >1.4, less than −1.4; p < 0.05, unpaired t test) in the Unc119−/− retina. The predicted upstream factors with an activation Z score >3.0 are shown. G, H, IPA networks showing interferon regulatory factor 1 (IRF1; G) and RELA proto-Oncogene (H), NF-kB subunit (RELA) as upstream regulators. IRF1 and RELA were predicted to be activated in the Unc119−/− mouse retina.
Inhibition of the JAK-STAT and NF-κB pathways ameliorates cone degeneration by Unc119 deficiency
Based on the results of RNA-seq analysis and IPA, we hypothesized that suppression of the JAK-STAT and NF-κB pathways may ameliorate cone degeneration in the Unc119−/− mouse retina. We injected curcumin, an inhibitor of the JAK-STAT and NF-κB pathways (Wang et al., 2018; Ashrafizadeh et al., 2020), into Unc119−/− mice from P21 for 7 d. At P28, we performed ERGs and histological analyses of control (DMSO-injected) Unc119−/− mice or curcumin-injected Unc119−/− mice (Fig. 6A). In the ERGs, we observed no significant differences in the amplitudes of scotopic a-, b-, and photopic a-waves. However, the amplitude of the photopic b-wave significantly increased in curcumin-treated mice, suggesting that curcumin partially rescued the decreased photopic b-wave amplitude observed in Unc119−/− mice at 1M (Fig. 6B–G). To examine whether curcumin suppresses the JAK-STAT and NF-κB pathways, we performed immunohistochemistry using an anti-Socs3 antibody for the JAK-STAT signature and an anti-RelA antibody for the NF-κB signature. Socs3 acts as a negative feedback regulator of the JAK-STAT pathway (Carow and Rottenberg, 2014; Hu et al., 2021), and RelA is one of the five NF-κB proteins (Liu et al., 2017). The signal intensities of Socs3 and RelA decreased in the curcumin-treated Unc119−/− retinas compared with those in the untreated Unc119−/− retinas, indicating that curcumin inhibits the JAK-STAT and NF-κB pathways (Fig. 6H,I). Next, we performed immunohistochemistry of the retinas of control and curcumin-injected Unc119−/− mice using anti-GFAP and anti-C1q antibodies. We measured the expression levels of GFAP and C1q in the retinas of the control and curcumin-injected mice to examine the effect of curcumin on inflammation. We observed a significant decrease in GFAP and C1q expression in the retinas of curcumin-treated mice, suggesting that curcumin suppressed inflammation resulting from Unc119 deficiency (Fig. 6J,K). To assess the cone properties, we immunostained control and curcumin-injected Unc119−/− mouse retinas for M-opsin and S-opsin. The number of M-opsin and S-opsin signals was not significantly affected in the curcumin-injected mouse retinas. However, mislocalization of M-opsin and S-opsin signals in the OPL was observed in the retinal sections from control mice, but not significantly detected in those from curcumin-treated mice (Fig. 6L,M). These results suggest that curcumin can ameliorate cone degeneration resulted from Unc119 deficiency by suppressing inflammation in the retina through the downregulation of JAK-STAT and NF-κB pathway activities.
Curcumin suppresses cone degeneration and inflammation in the Unc119−/− retina. A, Experimental design for curcumin administration to Unc119−/− mice. B–G, ERG analysis of curcumin-treated Unc119−/− mice at 1M. ERGs were recorded from control (DMSO) and curcumin-treated Unc119−/− mice at 1M (n = 4 per each genotype). B, Representative scotopic ERGs elicited by four stimulus intensities (−4.0 to 1.0 log cd s/m2) from control (DMSO) and curcumin-treated Unc119−/− mice at 1M. C, D, The scotopic amplitudes of a- (C) and b-waves (D) are shown as a function of the stimulus intensity. Data are presented as the mean ± SD. n = 4 per each genotype. n.s., not significant (two-way repeated-measures ANOVA, multiple comparisons). E, Representative photopic ERGs elicited by four stimulus intensities (−0.5 to 1.0 log cd s/m2) from control (DMSO) and curcumin-treated Unc119−/− mice at 1M. F, G, Photopic amplitudes of a- (F) and b-waves (G) are shown as functions of stimulus intensity. Data are presented as the mean ± SD. n = 4 mice per each genotype. *p < 0.05, n.s., not significant (two-way repeated-measures ANOVA, multiple comparison). H–M, Immunofluorescence analysis of retinal sections from control and curcumin-injected Unc119−/− mice was performed using antibodies against Socs3 (H), RelA (I), GFAP (J), C1q (K), M-opsin (L), and S-opsin (M). Nuclei were stained with DAPI (blue). The intensities of the Socs3, RelA, GFAP, and C1q signals in the retina were measured (H–K). The number of M-opsin- and S-opsin-positive cells was counted. The signal intensities of M-opsin and S-opsin in the OPL were measured. Arrowheads indicate mislocalization of M-opsin and S-opsin signals (L, M). Data are presented as mean ± SD. n = 4 per each genotype. *p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant (unpaired t test). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; OS, outer segment.
Inhibition of UNC119 binding to GNAT1 or GNAT2 by UNC119-K57X CRD mutant
Finally, to investigate the pathological mechanisms underlying CRD associated with a human UNC119 (hUNC119) mutation, we focused on a CRD mutation with a heterozygous A-to-T transition in codon 57, resulting in a lysine-to-premature termination codon (hUNC119-K57X) in the middle region of the hUNC119 protein (Kobayashi et al., 2000). We hypothesized that the hUNC119-K57X mutant would inhibit the interaction between hUNC119 and human GNAT1 (hGNAT1) or human GNAT2 (hGNAT2). To test this hypothesis, we performed immunoprecipitation analysis using plasmids expressing HA-tagged hUNC119, FLAG-tagged hGNAT1, FLAG-tagged hGNAT2, and Myc-tagged hUNC119-K57X. First, we examined the interaction between hUNC119 and hGNAT1 or hGNAT2 using immunoprecipitation and confirmed that hUNC119 interacts with both hGNAT1 and hGNAT2 (Fig. 7A). Next, we conducted an immunoprecipitation assay to determine whether the hUNC119-K57X mutant inhibited the interaction between hUNC119 and hGNAT1 or hGNAT2. The interaction between hUNC119 and hGNAT1 or hGNAT2 was inhibited in a dose-dependent manner by increasing the expression of the hUNC119-K57X mutant (Fig. 7B,C). We performed pull-down assays using GST and GST-fused hUNC119-K57X proteins with Nrl+/+ and Nrl−/− retinal lysates, followed by Western blotting for Gnat1, Gnat2, and Unc119. We did not observe an interaction between the hUNC119-K57X and Gnat1, Gnat2, or Unc119, suggesting that hUNC119-K57X indirectly inhibits the interaction between hUNC119 and hGNAT1 or hGNAT2 (Fig. S7). These results suggest that the hUNC119-K57X mutation competitively inhibits the functions of hGNAT1 and hGANT2, leading to the onset of CRD in a heterozygous hUNC119-K57X patient.
Inhibition of UNC119 interaction with GNAT1 or GNAT2 by UNC119-K57X. A, Immunoprecipitation analysis of human UNC119 (hUNC119) with hGNAT1 or hGNAT2. Plasmids expressing FLAG-tagged hGNAT1 or hGNAT2 and HA-tagged hUNC119 were cotransfected into HEK293T cells. The cell lysates were subjected to immunoprecipitation with the anti-FLAG antibody. Immunoprecipitated proteins were detected by Western blot analysis with anti-FLAG and anti-HA antibodies. B, C, Myc-tagged hUNC119-K57X, in addition to the expression constructs used in A, were cotransfected into HEK293T cells. The cell lysates were subjected to immunoprecipitation with the anti-FLAG antibody. Immunoprecipitated proteins were detected by Western blot analysis with anti-FLAG and anti-HA antibodies. Relative intensities were determined by quantification of IP: HA band intensities (normalized to Input: HA) using ImageJ software. D, A hypothetical model of CRD caused by compromised Unc119 function.
Discussion
In the current study, we observed retinal degeneration in Unc119−/− mice, resembling the pathology of CRD. Unc119 interacted with Gnat2 in cones, and Gnat2 localization was reduced in the Unc119−/− cone OS. RNA-seq results indicated a greater inflammatory state in Unc119−/− retinas. Curcumin administration suppressed inflammation and cone degeneration in Unc119−/− retinas. The hUNC119-K57X mutation associated with CRD inhibited the interaction between hUNC119 and hGNAT1 or hGNAT2, suggesting that the hUNC119-K57X mutation may competitively inhibit the function of hGNAT1 or hGNAT2. This study may advance our understanding of the pathological mechanisms underlying CRD and contribute to the development of novel therapeutic approaches.
Cone dysfunction followed by rod dysfunction in Unc119−/− mice
Electrophysiological and histological analyses of Unc119-deficient mice from 1M to 12M revealed pathological phenotypes similar to CRD (Figs. 2, 3; Figs. S3, S8). Cone dysfunction appeared at 1M, followed by progressive dysfunction and degeneration of cones and rods at 6M. Given that Unc119 expression was enriched in photoreceptor cells, we focused our analysis on photoreceptor degeneration. We examined the ERG b-wave implicit time and observed no significant differences between Unc119+/+ and Unc119−/− mice at 1M, 6M, or 12M (Figs. 2G,H, 3G,H; Fig. S3G,H), suggesting that Unc119 does not play an essential role in synaptic transmission from photoreceptors to bipolar cells, while recent studies have reported that Unc119 deletion reduces the steady-state glutamate release rate at rod synapses (Fehlhaber et al., 2023). Future studies on rod synapse in Unc119−/− mice might provide additional insights into a role of Unc119 in rod synapse architecture and function.
Unc119 interacts with Gnat2 in cones
Unc119 is known to interact with Gnat1 (Zhang et al., 2011) and the present study demonstrated that it also interacts with Gnat2 (Fig. 4C,D). Our pull-down and LC-MS/MS analysis results suggest that Ytdc2, Tba1a, and Hmmr also interact with Unc119 (Fig. 4B,C). Hmmr is a receptor for hyaluronic acid and plays an important role in cell migration, growth, and differentiation (Turley and Naor, 2012). In the eye, Hmmr is localized to the apical region of the retinal pigment epithelium and hyaluronan is a major component of the interphotoreceptor matrix (Hollyfield, 1999). Hmmr may be involved in photoreceptor survival. Examination of the possible functional interaction between Unc119 and Hmmr requires future analysis.
Additionally, in mice lacking the Unc119 gene, we observed reduced Gnat2 signals in photoreceptor OS by immunohistochemistry; however, the total amount of Gnat2 protein in the retina remained unchanged (Fig. 4E–G). This result suggests that Gnat2 does not localize properly to the photoreceptor OS in Unc119-deficient mice and that Unc119 may regulate Gnat2 transport in cones. It has been reported that a decrease in photopic ERGs responses without retinal degeneration occurs in adult mice lacking the Gnat2 gene (Chang et al., 2006), suggesting that the reduced localization of Gnat2 in the OS may contribute to cone photoreceptor dysfunction in Unc119−/− retinas.
Unc119 and regulation of G-protein α-subunit localization
The present study showed that Unc119 interacts with Gnat2 and is essential for its proper localization of Gnat2 in cones (Fig. 4E,F). Gnat1 is similarly regulated by Unc119 for its localization in rods (Zhang et al., 2011). In C. elegans, Unc119 regulates the proper localization of ODR-3 and GPA-13, G-protein alpha subunits, in the cilia of olfactory sensory neurons (Zhang et al., 2011). Based on these findings, we speculate that Unc119 may interact with a G-protein α-subunit and regulate its localization in cells other than retinal photoreceptor cells. For example, Gustducin and Gαolf, which are involved in taste and olfaction, respectively, are G-protein α-subunits that possess glycine at the second amino acid residue in each protein. Further studies are needed to examine whether Unc119 interacts with Gustducin and/or Gαolf in terms of taste and olfactory function. We observed no apparent changes in Unc119−/ mouse tissues except the retina as far as we examined. It should be noted that mouse Unc119 has a paralog, Unc119b, which might compensate for Unc119 function. Generating and analyzing Unc119−/−/Unc119b−/− double knock-out mice may provide additional insights into the function of Unc119 in various tissues and cells.
Unc119 deficiency induces immune and inflammatory responses in the retina
Our RNA-seq analysis identified 80 downregulated and 286 upregulated genes in the Unc119−/− mouse retina compared with the control retina (Fig. 5A). Initially, we focused on GFAP as one of the upregulated genes. Although the expression level of GFAP is typically low in Müller glial cells, it is highly expressed in the retina with photoreceptor degeneration (Sarthy et al., 1991), suggesting that the increased expression of GFAP is linked to photoreceptor degeneration caused by Unc119 deficiency. C1q, the first component of the classical complement pathway, is associated with pathological features of age-related neurodegenerative diseases such as amyloid-β accumulation in Alzheimer's disease, glaucomatous damage in glaucoma models, and drusen deposition in age-related macular degeneration (AMD; Howell et al., 2011; Dejanovic et al., 2022; Yednock et al., 2022; Greferath et al., 2024). Furthermore, the classical complement pathway, mediated by C1q, has been shown to contribute to the progression of retinal degeneration (Taylor et al., 2016; Jiao et al., 2018). Thus, elevated C1q expression may play a role in retinal degeneration in Unc119−/− retinas. Edn2 and Fgf2, genes commonly upregulated in retinal degeneration models (Chen et al., 2004; Rattner et al., 2008; Kuny et al., 2012; Samardzija et al., 2012; Bramall et al., 2013), were also upregulated in the Unc119−/− retina. Rdcvf2, an inactive thioredoxin secreted from rods, is equivalent to Rdcvf, which protects cones from degeneration (Chalmel et al., 2007). Thus, reduced expression of Rdcvf2 in the Unc119−/− retina may make cones more susceptible to degeneration. In addition, GSEA and IPA revealed elevated immune and inflammatory responses in Unc119−/− retinas (Fig. 5E,F). IPA further identified the activated transcription factors involved in the JAK-STAT and NF-κB signaling pathways in the Unc119−/− retina. This finding suggests that Unc119 deficiency activates these pathways and promotes photoreceptor degeneration. Curcumin treatment downregulated GFAP and C1q expression and reduced the ectopic localization of M-opsin and S-opsin in Unc119−/− mouse retinas (Fig. 6), suggesting that curcumin treatment suppressed cone photoreceptor degeneration caused by Unc119 deficiency. Overall, these results suggest that inhibition of the JAK-STAT and NF-κB signaling pathways ameliorates cone photoreceptor degeneration due to Unc119 deficiency.
Pathogenesis of CRD with UNC119-K57X mutation
The present study elucidated a part of the pathological mechanism underlying CRD associated with a C-terminal region-defective hUNC119 mutant, in which the codon encoding the 57th amino acid becomes a termination codon (Kobayashi et al., 2000). The K57X mutation decreased the interaction between UNC119 and GNAT1 or GNAT2 in a dose-dependent manner in the immunoprecipitation assay (Fig. 7B,C). Mutations in the GNAT1 gene have been reported to be associated with late-onset retinitis pigmentosa in humans (Carrigan et al., 2016). Gnat1-deficient mice show mild retinal degeneration with age (Calvert et al., 2000). GNAT2 mutations are well known to cause color blindness in humans (Hassall et al., 2017). Gnat2-deficient mice exhibit decreased cone-mediated responses during adulthood (Chang et al., 2006). Our findings suggest that the hUNC119-K57X mutant competitively inhibits the interaction between UNC119 and GNAT1 or GNAT2, leading to partial loss of function of GNAT1 or GNAT2, which may result in retinal degeneration. We propose that UNC119-K57X-mediated disruption of the interaction between UNC119 and GNAT1 or GNAT2 contributes to CRD pathogenesis. Previous research has shown that transgenic mice overexpressing Unc119-K57X exhibited a reduced b-wave amplitude in ERGs and thinning of the photoreceptor layer (Kobayashi et al., 2000). Further studies that generate and analyze heterozygous Unc119-K57X knock-in mice may be useful to confirm the pathological mechanism in vivo. Together, we propose a possible mechanism of CRD associated with UNC119 mutations, where the reduction in UNC119 function leads to decreased activity of both GNAT1 and GNAT2, accompanied by retinal inflammation, ultimately resulting in photoreceptor degeneration (Fig. 7D).
Footnotes
We thank U. Nakagawa for LC-MS/MS analysis; M. Kadowaki, M. Wakabayashi, Y Kinooka, N. Nakashima, and T. Kaneko for technical assistance; and Editage for English language editing. This work was supported by Grant-in-Aid for Scientific Research (21H02657, 24K09996) and Grant-in-Aid for Challenging Research (Exploratory; 23K18199) from the Japan Society for the Promotion of Science, AMED-CREST (21gm1510006) from the Japan Agency for Medical Research and Development, Japan Science and Technology Agency (JST) Moonshot R&D (JPMJMS2024), JST COI-NEXT (JPMJPF2018), OU Master Plan Implementation Project, and the Takeda Science Foundation.
The authors declare no competing financial interests.
This paper contains supplemental material available at: https://doi.org/10.1523/JNEUROSCI.2245-24.2025
- Correspondence should be addressed to Takahisa Furukawa at takahisa.furukawa{at}protein.osaka-u.ac.jp.
