Early proteomic and metabolic signatures of liver and eye in OAT-deficient mice

Abstract

Ornithine aminotransferase (OAT) deficiency causes hyperornithinemia and gyrate atrophy (GA) of the choroid and retina, a rare inherited retinal degeneration. To understand the early molecular changes that make the eye susceptible to damage, we performed quantitative proteomic and metabolomic profiling of liver, retina, and retinal pigment epithelium and choroid (RPE/Cho) from OAT-deficient (Oat^rhg) mice prior to detectable vision impairment. In addition to reduced OAT expression and elevated ornithine, methylation-related metabolites such as N(6)-methyl-lysine were altered in all examined tissues of Oat^rhg mice. In the liver, excess ornithine was directed into urea cycle metabolism, together with altered expression of detoxification enzymes and histone H2B proteins. In contrast, the retina showed minimal proteomic changes but pronounced alterations in amino acid pathways that support glutamate homeostasis. The RPE/Cho demonstrated the most extensive proteomic changes, particularly in mitochondrial metabolism, cytoskeleton, and extracellular matrix, along with changes in metabolites involved in lysine metabolism, energy metabolism, and antioxidant capacity. Incubation with ¹³C lysine demonstrated that lysine was primarily degraded in RPE/Cho but not the retina, and ornithine enhanced lysine degradation in an OAT-dependent manner. Together, these findings highlight common and tissue-specific impacts of OAT on the liver and ocular tissues and provide insight into early molecular changes that contribute to the selective vulnerability of the eye in GA. Proteomics data are available via ProteomeXchange (PXD063614) and metabolomics data via MassIVE repository (MSV000101103).

1. Introduction

Ornithine aminotransferase (OAT) is a pyridoxal-5-phosphate (PLP)-dependent mitochondrial enzyme that catalyzes the reversible transamination of ornithine and α-ketoglutarate (α-KG) to generate glutamate-5-semialdehyde (GSA) and glutamate. GSA spontaneously cyclizes to pyrroline-5-carboxylate (P5C), a precursor for proline and glutamate synthesis. As ornithine is a key urea cycle intermediates, OAT connects both the urea cycle and the tricarboxylic acid (TCA) cycle (Fig. 1A). Mutations in the OAT gene cause gyrate atrophy (GA) of the choroid and retina, a rare autosomal recessive blinding disorder characterized by markedly elevated plasma ornithine levels (Takki, 1974; Wang et al., 1996; Valle et al., 1981a). The clinical presentation of GA is dominated by ocular pathology, with rare and mild neurological and muscular manifestations (Peltola et al., 2002; Sipilä et al., 1979; Fleury et al., 2007). Ocular symptoms typically present in childhood with myopia and night blindness, followed by progressive chorioretinal atrophy and peripheral vision loss. Cataracts and macular edema may also occur, with variable rate of progression (Balfoort et al., 2024, 2025; Kaiser-Kupfer et al., 1985; Sen et al., 2021).

Fig. 1. Experimental workflow and mouse model validation.

(A) Illustration of the OAT reaction and connected metabolic pathways. (B) Schematic overview of the experimental workflow. (C) Western blot showing near-absent OAT protein expression in retina, RPE/Cho, and liver tissues in Oat^rhg compared to Oat^{rhg ±} and (D) quantification normalized to GAPDH loading control (N = 3). p < 0.05. RPE/Choroid (RPE/Cho); α-ketoglutarate (α-KG); Pyrroline-5-carboxylate (P5C); Ornithine aminotransferase (OAT); Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Tricarboxylic acid cycle (TCA cycle).

A distinctive biochemical feature of GA in adult patients is a 10- to 20-fold elevation of plasma ornithine. High ornithine has long been considered toxic to the eye; however, the underlying mechanisms remain poorly understood (Simell and Takki, 1973; Valle et al., 1981b; Kaiser-Kupfer et al., 2004; Ueda et al., 1998). Importantly, plasma ornithine levels do not consistently correlate with disease severity, and hyperornithinemia alone, as observed in hyperornithinemia-hyperammonemia-homocitrullinuria (HHH) syndrome, does not affect vision (Jasani et al., 2018; Hommes et al., 1986). In addition to ornithine, levels of multiple plasma amino acids are altered in GA patients, including lysine, creatine, proline, and histidine (Balfoort et al., 2025; Valle et al., 1981b; Palmer et al., 2023). Plasma lysine is consistently decreased, likely due to the competition with ornithine and arginine for shared renal transporters, as lysinuria has been reported in some patients (Valle et al., 1981b; Peltola et al., 2000; Khan et al., 1994). Moreover, plasma creatine is reduced, reflecting the inhibition of arginine:glycine amidinotransferase (AGAT) by accumulated ornithine (Valayannopoulos et al., 2009). This reduction may contribute to extraocular manifestations in GA, particularly affecting muscles and the brain (Valayannopoulos et al., 2009; Nänto-Salonen et al., 1999; Sipilä et al., 1980, 1981). In addition, ornithine can be funneled through ornithine decarboxylase (ODC) into polyamine synthesis, a process that consumes S-adenosylmethionine (SAM), a universal methyl donor for the methylation of DNA, histones, and small molecules (Bistulfi et al., 2009; Pirini et al., 2025). SAM links methylation to both redox and nucleotide metabolism by generating homocysteine for glutathione synthesis and interacting with the folate cycle that supports purine production. Elevated polyamines have been shown to be toxic to retinal pigment epithelium (RPE) cells, induce RPE dysfunction (Kaneko et al., 2007; Ohashi et al., 2017), and increased polyamine levels have been reported in plasma and urine of some GA patients.

Currently, there is no treatment for GA. Disease management aims to reduce systemic ornithine with an arginine-restricted diet and support downstream pathways using supplements such as proline, creatine, lysine, and pyridoxine (vitamin B6) (Sipilä et al., 1981; Kim et al., 2013; Hayasaka et al., 1985; Kennaway et al., 1980; Ohkubo et al., 2005; Elpeleg and Korman, 2001; Zekušić et al., 2018). These interventions can slow disease progression; however, most patients ultimately become blind. Recent work has shown that liver-targeted gene therapy reduces systemic ornithine and partially improves retinal function and structure in OAT-deficient mice (Boffa et al., 2023). Importantly, combined restoration of OAT in both liver and retina provides greater improvements in retinal structure and function than targeting either tissue alone (Dell’Aquila et al., 2025), supporting that the loss of OAT activity in both tissues contributes to the disease phenotype. However, how OAT deficiency and hyperornithinemia affect these tissues at the molecular level, particularly in driving vision impairment, remains undefined. To identify early molecular changes associated with OAT deficiency prior to overt retinal degeneration, we performed integrated proteomic and metabolomic profiling of liver, retina, and RPE/choroid (RPE/Cho) from young adult Oat^rhg mice (Oat^G353A/G353A), which recapitulate the key biochemical and ocular features of GA (Bisaillon et al., 2014). We found that OAT deficiency triggers tissue-specific responses, affecting amino acid metabolism, the urea cycle, SAM-dependent pathways, detoxification, and remodeling of histone and extracellular matrix (ECM) proteins. These findings provide a mechanistic framework of how OAT deficiency disrupts cellular homeostasis in the liver and eye.

2. Materials and methods

2.1. Animals

All animal procedures were conducted in accordance with National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee (IACUC) of West Virginia University. All animals were housed under standard conditions with ad libitum access to standard chow and water. Oat homozygous mutants (Oat^rhg) mice were obtained from the Jackson Laboratory and bred onto a C57BL/6J background for more than three generations before use in experiments. Heterozygous Oat^{rhg ±} mice, which retain OAT activity comparable to wild-type controls, were used as controls in this study (Bisaillon et al., 2014). Adults at 3.5 months of age from both genotypes were used for experiments. Based on prior reports, retinal and RPE/Choroidal (RPE/Cho) structures and functions are normal at these ages, allowing us to analyze molecular changes preceding visual dysfunction (Boffa et al., 2023; Bisaillon et al., 2014). Genotyping was performed by Transnetyx, Inc. (Cordova, TN; see Table S1).

2.2. Tissue collection

Samples for metabolomics and proteomics were collected around noon from separate sets of adult Oat^rhg and littermate Oat^{rhg ±} control mice of both sexes. Prior to tissue collection, all animals were transferred to a procedure room and acclimated for 1 h. Mice were euthanized via cervical dislocation. Their eyes were immediately enucleated, and the surrounding muscle and fat were carefully removed. The neural retina and RPE/Cho were then isolated in cold Hank's Balanced Salt Solution (HBSS) by dissecting the anterior segments as previously described (Zhu et al., 2018). Because the RPE is difficult to separate from its tightly bound choroid, and the choroid plays an important role in RPE function, we analyzed the RPE/Cho complex as in our previous studies (Xu et al., 2020). A small section from the right lobe of the liver was excised and rapidly rinsed in cold Phosphate-Buffered Saline (PBS). All collected tissues were flash-frozen in liquid nitrogen.

2.3. Proteomics sample preparation

Proteins from the retina, RPE/Cho, and liver were extracted using radioimmunoprecipitation assay (RIPA) buffer with protease and phosphatase inhibitors (5 mg/mL). The tissue lysates were centrifuged at 12,000 rpm for 15 min at 4°C, and the supernatant was used to measure protein concentration by BCA assay. The quantitative proteomic analysis was performed by IDeA National Resource for Quantitative Proteomics (Little Rock, Arkansas).

2.4. Quantitative proteomics

Total protein from each sample was reduced, alkylated, and purified by chloroform/methanol extraction prior to digestion with sequencing-grade modified porcine trypsin (Promega). Tryptic peptides were then separated by reverse phase XSelect CSH C18 2.5 μm resin (Waters) on an in-line 150 × 0.075 mm column using an UltiMate 3000 RSLCnano system (Thermo). Peptides were eluted using a 60 min gradient from 98:2 to 65:35 buffer A:B ratio (Buffer A contains 0.1% formic acid, 0.5% acetonitrile; Buffer B contains 0.1% formic acid, 99.9% acetonitrile). Eluted peptides were ionized by electrospray (2.2 kV) followed by mass spectrometric analysis on an Orbitrap Exploris 480 mass spectrometer (Thermo). To assemble a chromatogram library, six gas-phase fractions were acquired on the Orbitrap Exploris with 4 m/z DIA spectra (4 m/z precursor isolation windows at 30,000 resolution, normalized AGC target 100%, maximum inject time 66 ms) using a staggered window pattern from narrow mass ranges using optimized window placements. Precursor spectra were acquired after each DIA duty cycle, spanning the m/z range of the gas-phase fraction (i.e. 496-602 m/z, 60,000 resolution, normalized AGC target 100%, maximum injection time 50 ms). For wide-window acquisitions, the Orbitrap Exploris was configured to acquire a precursor scan (385-1015 m/z, 60,000 resolution, normalized AGC target 100%, maximum injection time 50 ms) followed by 50 × 12 m/z DIA spectra (12 m/z precursor isolation windows at 15,000 resolution, normalized AGC target 100%, maximum injection time 33 ms) using a staggered window pattern with optimized window placements. Precursor spectra were acquired after each DIA duty cycle.

2.5. Proteomics data analysis

Following data acquisition, data were searched using an empirically corrected library against the UniProt Mus musculus database (Proteome ID: UP000000589, 2nd version of 2023), and a quantitative analysis was performed to obtain a comprehensive proteomic profile. Proteins were identified and quantified using EncyclopeDIA (Searle et al., 2018) and visualized with Scaffold DIA using 1% false discovery thresholds at both the protein and peptide level. Protein MS2 exclusive intensity values were assessed for quality using ProteiNorm (Graw et al., 2020). The data were normalized using Cyclic Loess (Bolstad et al., 2003) to correct for systemic variation. Differential expression was assessed using proteoDA, which implements linear models (limma) with empirical Bayes (eBayes) with empirical Bayes (eBayes) smoothing to the standard errors (Ritchie et al., 2015; Thurman et al., 2023). Proteins with an adjusted p-value <0.05 and a fold change (FC) > 2 were considered significant.

2.6. Proteomic pathway analysis

Pathway enrichment and functional annotation of differentially expressed proteins were performed using STRING (v12.0) (Szklarczyk et al., 2023) and DAVID (Sherman et al., 2022). Statistical significance was defined as ∣log2fold change∣ ≥ 1 and a p-value <0.05. Venn diagrams were generated using InteractiVenn (Heberle et al., 2015) to visualize protein overlaps across tissues. Metabolic proteins were identified by intersecting significantly altered proteins with a previously published metabolic gene set (Li et al., 2020), followed by manual curation using UniProt and NCBI/PubMed. For each protein, literature validation was performed using Boolean query strings combining the protein identifier with keywords for ocular tissues, liver, and metabolic function (e.g., “[protein name] AND (retina OR RPE OR liver OR metabolism)”).

2.7. Retina and RPE/Cho explant culture

Mouse retina and RPE/Cho were quickly isolated in 200 μL of cold HBSS as previously described (Xu et al., 2020; Yam et al., 2019). Tissues were then transferred to 200 μL of pre-equilibrated Krebs-Ringer buffer (KRB) containing ¹³C₆-lysine, with or without unlabeled ornithine. All experimental media were supplemented with 5 mM glucose and incubated at 37°C in a CO₂ incubator for up to 2 hours. Following incubation, tissues were collected, snap-frozen in liquid nitrogen, and stored at −80°C until analysis.

2.8. Metabolite analysis and data processing

Metabolites were extracted from tissues by homogenization in ice-cold 80% (v/v) methanol and from plasma in ice-cold 100% methanol, as previously described (Du et al., 2015). Homogenates were centrifuged at 13,000 rpm for 15 min at 4°C, and supernatants were vacuum-dried (Eppendorf Vacufuge Plus) and analyzed in liquid chromatography-mass spectrometry (LC-MS). Targeted metabolite profiling was performed on a Shimadzu Nexera UHPLC system coupled to an AB Sciex QTRAP 5500 mass spectrometer (AB Sciex, Toronto, ON, Canada) using a multiple reaction monitoring (MRM) method as previously described (Li et al., 2020; Saravanan et al., 2023). Metabolite peaks were identified and quantified using MultiQuant Software (v3.0). Statistical analysis, volcano plots (thresholds: P < 0.05 and fold change >1.3), Partial Least Squares Discriminant Analysis (PLS-DA), and pathway enrichment analysis were performed using MetaboAnalyst 6.0 (https://www.metaboanalyst.ca). Raw mass spectrometry data have been deposited to MassIVE repository (MSV000101103).

2.9. Western blot

Immunoblots were performed as previously described (Zhu et al., 2022). Briefly, tissue samples containing 20 μg protein were separated in Mini-PROTEAN TGX precast gels, transferred onto 0.22 μm nitrocellulose membranes, and blocked with 5% non-fat milk in 1x Tris-Buffered Saline with 0.1% Tween-20 (TBST). Membranes were incubated overnight at 4°C with primary antibodies (1:1000 dilution; see Table S1) and then in horseradish peroxidase (HRP)-conjugated secondary antibodies (1:2000) for 1 h, followed by visualization by chemiluminescence.

2.10. Statistics

Ornithine ion abundances in Oat^rhg mice were divided by those from Oat^{rhg ±} mice to calculate fold changes (FC). Data were graphed using GraphPad Prism (v9.5.1). Statistical significance was determined using multiple unpaired t-tests with Bonferroni correction method, with p < 0.05 considered significant. All data are presented as the mean ± SD.

3. Results

3.1. OAT deficiency causes ocular and liver ornithine accumulation before visual decline in Oat^rhg mice

To capture early molecular changes that precede overt degeneration, we performed quantitative proteomics and targeted metabolomics on the retina, RPE/Cho, and liver from 3.5-month-old Oat^rhg mice and Oat^{rhg ±} littermate controls (Boffa et al., 2023; Ginguay et al., 2017) (Fig. 1B). At this age, Oat^rhg mice appear healthy, with body size comparable to Oat^{rhg ±} controls. Although loss of OAT enzymatic activity in Oat^rhg mice has been established (Bisaillon et al., 2014), OAT protein expression in ocular tissues has not been assessed. Western blot analysis demonstrated a marked reduction of OAT protein expression in retina, RPE/Cho, and liver of Oat^rhg mice (Fig. 1C, D, S1). Consistent with previous reports, plasma ornithine levels were more than 6-fold higher in adult Oat^rhg mice (Cleary et al., 2005; de Sain-van der Velden et al., 2012; Wang et al., 1995). Ornithine ion abundance in the liver, retina, and RPE/Cho was also dramatically increased, exceeding corresponding control tissue by more than 10-fold (Fig. S2). Despite systemic and tissue ornithine accumulation, electroretinogram (ERG) recordings showed normal visual function in 3.5-month-old Oat^rhg mice, indicating that retinal degeneration had not yet developed at this stage (Fig. S3). These data confirm global ornithine accumulation and lack of OAT protein in ocular and liver tissues, establishing this pre-degenerative stage as suitable for multi-omics analyses to define early molecular changes that may underlie the later selective vulnerability of the eye.

3.2. OAT deficiency reprograms liver proteome beyond ornithine metabolism

To determine the molecular changes underlying OAT deficiency in the liver, we conducted quantitative proteomic profiling of tissues from adult Oat^rhg and Oat^{rhg ±} mice. The proteomics analysis identified 62 differentially expressed (DE) proteins in adult Oat^rhg mouse liver compared to controls (Fig. 2A). As expected, OAT was the most significantly changed protein with an approximately 9-fold decrease. Functional categorization of these DE proteins revealed a significant enrichment in metabolism (34%), followed by transcription (18%), cytoskeleton (13%), immunity (11%), and signal transduction (10%) (Fig. 2B). Subcellular localization analysis showed that these proteins were predominantly associated with the cytoplasm, endoplasmic reticulum (ER), nucleus, and membranes, while mitochondrial and lysosomal proteins were less affected (Fig. 2C). Gene ontology (GO) analysis revealed that the top five enriched biological pathways in the OAT-deficient liver were nucleosome assembly, steroid, alcohol, and xenobiotic metabolism, as well as oxidative demethylation (Fig. 2D).

Fig. 2. Analysis of DE proteins in Oat^rhg liver tissues compared to controls.

(A) Volcano plot showing top 5 up and downregulated proteins in Oat^rhg liver and (B) their biological function classification based on manual search in UniProt and NCBI/PubMed. (C) Subcellular localization of DE proteins in the liver based on UniProt annotations. (D) Top five enriched biological processes identified from the analysis of DE proteins using STRING. (E) Extended view of enriched pathways and their associated proteins. (F) Functional annotation analysis of enriched metabolic proteins in the liver. Detox-Phase I (oxidation/reduction by cytochrome P450 enzymes); Detox-Phase II (conjugation reactions such as sulfation and ester hydrolysis); Major Histocompatibility Complex (MHC). N = 3. ∣log₂fold change∣ ≥ 1, p-value <0.05.

To further define these changes, we examined key altered proteins within these categories (Fig. 2E). We found that several histone H2B variants (H2B1F, H2B1M, H2B1C) and demethylases (KDM3B) were upregulated, suggesting chromatin remodeling and transcriptional alterations. In addition, we observed substantial disruption of the Phase I and II detoxification pathways, involving cytochrome P450 enzymes (CP341, CP17A, CP3AG, CP341) and sulfotransferases (ST2A1, ST2A5, ST2A8, ST3A1), which are critical for the metabolism of drugs, steroids, bile acids, and fatty acids. Flavin-containing monooxygenase 3 (FMO3), a key hepatic monooxygenase, was also markedly reduced, potentially altering the detoxification capacity of the liver in Oat^rhg mice. To refine the metabolic changes in the liver, we next performed GO analysis on DE metabolic proteins, examining biological processes, cellular components, and molecular functions (Fig. 2F). The analysis showed that altered metabolic proteins in Oat^rhg liver were enriched in steroid and lipid metabolism, mainly localized to the ER and microsomes, and associated with oxidoreductase, monooxygenase, and transferase activities. Taken together, these results illustrate that OAT deficiency in the adult liver extends beyond amino acid transamination, potentially impacting diverse pathways including steroid and lipid metabolism, detoxification, and transcriptional regulation.

3.3. Metabolomic analysis reveals early alterations of metabolic processes in Oat^rhgliver

Targeted metabolomics quantified 144 metabolites in liver from Oat^rhg and Oat^{rhg ±} mice. The Partial Least Squares Discriminant Analysis (PLS-DA) showed clear separation between genotypes, indicating a distinct metabolic profile in Oat^rhg liver (Fig. 3A). Statistical comparison between groups identified 7 significantly altered metabolites, 3 increased and 4 decreased in Oat^rhg compared to Oat^rhg+/−, as shown in Fig. 3B. Consistent with the canonical role of OAT in ornithine catabolism, targeted metabolomics revealed marked accumulation of ornithine and citrulline in Oat^rhg liver, together with increased betaine and decreased glutathione, lactate, fructose, and N(6)-methyl-lysine. KEGG enrichment analysis highlighted the involvement of these metabolites in arginine and proline biosynthesis, one-carbon metabolism, glutathione metabolism, fructose and mannose metabolism, glycine/serine/threonine metabolism, and fatty acid degradation (Fig. 3C). Together, these results reveal that loss of OAT activity triggers a broad metabolic reprogramming in the liver, in which excess ornithine is redirected into the urea cycle and polyamine synthesis, with secondary consequences for energy metabolism, S-adenosyl methionine (SAM)-dependent methylation, and glutathione-mediated redox homeostasis (Fig. 3D).

Fig. 3. Metabolite analysis in liver tissues of Oat^rhg compared to control mice.

(A) Partial least squares discriminant analysis (PLS-DA) of liver metabolomics distinguishes Oat^rhg from Oat^{rhg ±} mice. (B) Volcano plot of liver metabolomics showing increased (right) and decreased (left) metabolites in Oat^rhg relative to Oat^rhg+/−. (C) KEGG pathway enrichment analysis of liver metabolomics comparing Oat^rhg and Oat^rhg+/−. (D) Proposed model of metabolic alterations in Oat^rhg liver. In the liver, excess ornithine is redirected into the urea cycle, increasing the demand for cytosolic aspartate and competing with the malate-aspartate shuttle. Fructosederived pyruvate provides an alternative carbon source to sustain both TCA cycle activity and shuttle function. Excess ornithine may also be utilized for polyamine synthesis, affecting the SAM availability and downstream pathways. Red indicates increased, blue decreased in Oat^rhg relative to controls. Darker arrows indicate putative enhanced metabolic flux. Pyruvate (Pyr); Aspartate (Asp); Oxaloacetate (OAA); Malate (Mal); Pyrroline-5-carboxylate (P5C); Ornithine aminotransferase (OAT); S-adenosyl methionine (SAM); Me-Lysine (N6-Methyl-Lysine). N = 4. Fold change ≥1.3, p-value <0.05.

3.4. OAT deficiency induces early disruption of retinal amino acid metabolism

Proteomic analysis of retina identified only 5 significantly DE proteins in Oat^rhg mice compared to controls at this early stage. Among these, 3 were downregulated, including OAT, and 2 upregulated (Fig. 4A). These proteins were primarily associated with cytoskeletal regulation (LIMA1, ROCK1) and synapse modulation (CAD13).

Fig. 4. Proteome and metabolome changes in Oat^rhg mouse retinas.

(A) Volcano plot showing top 5 DE proteins in Oat^rhg retina (N = 3, ∣log₂fold change∣ ≥ 1, p-value <0.05). (B) Partial least squares discriminant analysis (PLS-DA) of retina metabolomics distinguishes Oat^rhg from Oat^{rhg ±} mice. (C) Volcano plot showing significantly increased (right) and decreased (left) metabolites in Oat^rhg retina compared to Oat^rhg+/− (N = 4, fold change ≥1.3, p-value <0.05). (D) KEGG pathway enrichment analysis of retinal metabolomics comparing Oat^rhg and Oat^rhg+/−. (E) Proposed model of metabolic alterations in Oat^rhg retina. The retina is deprived of several amino acids supplied from the RPE. BCAAs transamination may provide nitrogen for glutamate synthesis. Elevated ornithine may also compete with other cationic amino acids for shared transporters. Red indicates increased and blue decreased metabolites in Oat^rhg relative to controls. Darker and dotted arrows indicate putative enhanced metabolic flux or transport. Branched chain amino acids (BCAAs); Glutamate (Glu); Pyrroline-5-carboxylate (P5C); Ornithine aminotransferase (OAT); S-adenosyl methionine (SAM); N1-Acetylspermidine (Ac-Spm); N1-Methyl-4-pyridone-3-carboxamide (4PY); Cystathionine (Cystathio); N(6)-Methyl-Lysine (Me-Lysine); L-type amino acid transporter 1/2 (LAT1/2); Cationic amino acid transporter 1 (CAT1); N = 4. Fold change ≥1.3, p-value <0.05.

PLS-DA confirmed clear separation between Oat^rhg and controls in retinal metabolome (Fig. 4B). Targeted metabolomics detected 126 metabolites in the retina, of which 11 were significantly altered in Oat^rhg compared to Oat^{rhg ±} mice (2 increased and 9 decreased) (Fig. 4C). Ornithine was increased by more than 19-fold, together with a modest increase in adipic acid, an intermediate of dicarboxylic fatty acid metabolism. In contrast, multiple amino acids, including lysine, arginine, valine, and leucine, were reduced, together with lysine-derived aminoadipic acid. Several metabolites linked to one-carbon and polyamine pathways were also decreased, including N(6)-methyl-lysine, N1-methyl-4-pyridone-3-carboxamide (4PY), cystathionine, xanthosine, and N1-acetylspermidine, indicating reduced availability of methyl donors and altered nucleotide and redox metabolism. KEGG analysis showed enrichment in glutathione metabolism and multiple amino acid pathways, including arginine and proline, lysine, BCAAs, and aromatic amino acid metabolism (Fig. 4D). Overall, despite minimal proteomic changes, metabolomics indicate that OAT deficiency alters retinal amino acid metabolism, suggesting an early metabolic adaptation prior to overt retinal dysfunction (Fig. 4E).

3.5. RPE/Cho proteomics reveal early mitochondrial and structural protein changes in OAT deficiency

Among the three analyzed tissues, the RPE/Cho showed the most pronounced early proteomic changes, with 254 DE proteins in Oat^rhg compared with Oat^{rhg ±} mice (155 downregulated and 99 upregulated; Fig. 5A). These alterations involved metabolism (38%), cytoskeleton (15%), and signaling pathways (13%), and a substantial upregulation of crystallin proteins (CRYAA, CRBB2) (Fig. 5A and B). GO analysis indicated reduced mRNA processing, ER-Golgi trafficking, and mitochondrial bioenergetics (i.e., electron transport chain). In contrast, pathways related to lipid and fatty acid metabolism, as well as cellular transport processes, were upregulated (Fig. 5C). Most of these DE proteins were localized to the cytoplasm, mitochondria, and ER (Fig. 5D). Strikingly, we found that the expression of several ribosomal subunit proteins (RL21, RL32, RL34, RS21) and the mitochondrial ribosomal protein RT35 were significantly decreased (Fig. 5E). The RPE has a robust mitochondrial metabolism that supports energy production and nutrient synthesis for local use and export to the retina (Adijanto et al., 2014; Tong et al., 2022; Zhu et al., 2023a). Several mitochondrial proteins, including mitochondrial transporters (TOM22, MPC1), components of the electron transport chain (CX7A1, COX2, NDUS8, NDUF4), amino acid (P5CR2, KCRS) and fatty acid metabolism enzymes (CPT1B, ABHDB, ACPM), were downregulated in RPE/Cho, suggesting an early disruption of mitochondrial function. In contrast, crystallin family proteins (CRYAA, CRBB1, CRBB2) were upregulated, potentially reflecting cellular stress response, consistent with previous reports (Markitantova and Simirskii, 2025; Kannan et al., 2016).

Fig. 5. Proteomic changes in RPE/Cho in Oat^rhg compared to controls.

(A) Volcano plot highlighting the top 5 up- and down-regulated proteins in Oat^rhg RPE/Cho. (B) Pie chart shows enriched biological processes in RPE/Cho. (C) GO analysis of DE proteins in Oat^rhg RPE/Cho showing enriched biological processes and (D) cellular components. Up and downregulated proteins were analyzed separately and are represented by red and blue bar graphs, respectively. (E) Heatmaps illustrate proteins associated with selected enriched biological and cellular processes. (F) List of DE proteins in Oat^rhg associated with thick filaments (myosins), intermediate filaments (keratins), and extracellular matrix (ECM). Red and blue indicate upregulated and downregulated proteins, respectively. Endoplasmic reticulum (ER); Electron transport chain (ETC). N = 3. ∣log₂fold change∣ ≥ 1, p-value <0.05.

Given the essential role of RPE metabolism in supporting retinal health (Zhu et al., 2023b; Hurley, 2021), we further analyzed DE metabolic proteins in the RPE/Cho. GO enrichment analysis showed that these proteins were mainly involved in lipid and fatty acid metabolism, the respiratory chain, carbohydrate metabolism, and transport-related processes (Fig. S4). The enriched proteins were predominantly localized to mitochondria, ER, microsomes, and inner mitochondrial membrane and were enriched for oxidoreductase, monooxygenase, hydrolase, and transferase activities. Interestingly, significant alterations were found in cytoskeletal, intermediate filament, and ECM proteins, with downregulation of myosins and upregulation of keratins and a few collagen proteins (Fig. 5F). Overall, loss of OAT activity in RPE/Cho is associated with early proteomic changes characterized by impaired mitochondrial bioenergetics, reduced ribosomal proteins, and cytoskeletal-ECM remodeling.

3.6. OAT deficiency alters amino acid metabolism and redox capacity in RPE/Cho

Targeted metabolomics quantified 136 metabolites in RPE/Cho from Oat^rhg and Oat^{rhg ±} mice. PLS-DA showed that Oat^rhg and control RPE/Cho formed distinct clusters, indicating that metabolite changes clearly separated the two groups (Fig. 6A). Statistical analysis identified 7 significantly altered metabolites, of which 2 were increased and 5 decreased in Oat^rhg RPE/Cho (Fig. 6B). Ornithine was increased by more than 20-fold, accompanied by a more than 2-fold elevation in aminoadipic acid. Several metabolites linked to SAM-dependent pathways, including creatinine, 4PY, and N(6)-methyl-lysine, were reduced, consistent with altered methylation pathways. Decreased carnosine and myo-inositol further suggest alteration of antioxidant defense and phosphoinositide-related pathways, which are important for membrane dynamics and cellular structure (De Craene et al., 2017; Hou et al., 2025; Lourdes et al., 2024; de Almeida Torres et al., 2023; Caruso et al., 2023a). KEGG enrichment analysis identified that OAT deficiency affects glutathione metabolism, arginine and proline metabolism, and the inositol phosphate pathway, as shown in Fig. 6C. These data indicate that OAT deficiency in RPE/Cho is associated with alteration of methylation, inositol-related processes, and antioxidant metabolites.

Fig. 6. Altered metabolites in RPE/Cho of Oat^rhg compared to controls.

(A) Partial least squares discriminant analysis (PLS-DA) of RPE/Cho metabolomics distinguishes Oat^rhg from Oat^{rhg ±} mice. (B) Volcano plot of RPE/Cho metabolomics showing increased (right) and decreased (left) in Oat^rhg relative to Oat^rhg+/−. (C) KEGG pathway enrichment analysis of RPE/Cho metabolomics comparing Oat^rhg and Oat^rhg+/−. (D) Proposed model of metabolomic alterations in Oat^rhg RPE/Cho. Excess ornithine may be redirected toward polyamine synthesis, impacting the SAM pool and downstream pathways. In addition, high ornithine may compete with other cationic amino acids for shared transporters, altering their cellular uptake. Red indicates increased and blue decreased metabolites in Oat^rhg relative to controls. Darker arrows indicate putative enhanced metabolic flux or transport. Pyrroline-5-carboxylate (P5C); ornithine aminotransferase (OAT); Glutamate (Glu); Lysine (Lys); Aminoadipic acid (2-AAA); Coenzyme A (CoA); S-adenosyl methionine (SAM); Me-Lysine (N6-Methyl-Lysine); N1-Methyl-4-pyridone-3-carboxamide (4PY); Histidine (His); β-Alanine (β-Ala); Cationic amino acid transporter 1 (CAT1); Phosphoinositol (PI); Photoreceptor outer segment (POS). N = 4. Fold change ≥1.3, p-value <0.05.

3.7. OAT deficiency alters lysine degradation in RPE/Cho but not the retina

To investigate how OAT and ornithine affect lysine metabolism, we incubated RPE/Cho and retina from Oat^rhg and control mice ex vivo in ¹³C-lysine with or without exogenous ornithine (Fig. 7A and B). Lysine is mainly degraded in mitochondria into acetyl-CoA for the TCA cycle, generating aminoadipic acid, glutaric acid, glutamate, and glutamine, and in peroxisomes, producing pipecolic acid. Endogenous lysine was rapidly replaced by ¹³C-lysine with similar isotopic enrichment among all groups (Fig. S5). Lysine degradation products were quickly labeled by ¹³C in RPE/Cho but not the retina in all conditions, suggesting lysine catabolism is confined to the RPE but absent in the retina (Fig. 7C). In RPE/Cho, ornithine substantially decreased ¹³C enrichment in lysine-derived intermediates; however, OAT deficiency reversed this decrease and further increased the enrichment of glutaric acid and glutamate (Fig. 7C). These data suggest that ornithine affects lysine catabolism in RPE/Cho in an OAT-dependent manner, and OAT deficiency enhances lysine degradation (Fig. 7D).

Fig. 7. 13C-lysine tracing in Oatrhg RPE/Cho.

(A) Schematic of lysine degradation pathway. (B) RPE/Cho and retina from Oat^{rhg ±} and Oat^rhg were incubated with 1 mM ¹³C-lysine and 0 or 1 mM ornithine in Krebs-Ringer bicarbonate buffer (KRB) for 2 h, followed by metabolite analysis. (C) ¹³C fractional enrichment of metabolites related to lysine catabolism in RPE/Cho and retina following incubation with ¹³C-lysine. Statistical comparisons were performed relative to Oat^{rhg ±} without ornithine (*p-value <0.05) and Oat^{rhg ±} with ornithine (#p-value <0.05). (D) Proposed model of OAT- and ornithine-dependent regulation of lysine metabolism in RPE/Cho and retina under control conditions (left) and OAT deficiency (right). Gradient arrows indicate potential export of lysine-derived metabolites from RPE/Cho to retina. Dotted arrows indicate limited lysine catabolism in retina, whereas thick arrows indicate enhanced lysine catabolism in RPE/Cho. Pyrroline-5-carboxylate (P5C); α-ketoglutarate (α-KG); Glutamate (Glu); Tricarboxylic acid cycle (TCA cycle); Coenzyme A (CoA). N > 2, p-value <0.05.

3.8. OAT deficiency drives tissue-specific proteomic remodeling with few shared changes across tissues

To investigate common effects of OAT deficiency in tissue proteomes, we compared DE proteins across RPE/Cho, retina, and liver (Fig. 8A). OAT was the only protein consistently reduced in all three tissues at both ages, confirming the robustness of our model. RPE/Cho showed the most extensive proteomic remodeling, followed by liver and then retina, supporting RPE/Cho as the initial site of damage in OAT deficiency. Aside from OAT, six proteins were consistently altered across all three tissues. Proteins such as TNNT3, CP341, BGAL, and Q8HWB2 changed in the same direction in each tissue, whereas MYH7B and CPLX4 showed opposite regulation between RPE/Cho and liver (Fig. 8B). These shared proteins are involved in cytoskeletal structure and vesicle trafficking (TNNT3, MYH7B, CPLX4), lysosomal processing (BGAL), immune signaling (Q8HWB2), and lipid metabolism (CP341), highlighting a common cellular response to OAT loss.

Fig. 8. Cross-tissue proteomic overview of DE proteins in Oat^rhg compared to control mice.

(A) Total significantly changed and overlapping proteins in Oat^rhg in the RPE/Cho, retina, and liver tissues. (B) Overlapping DE proteins with the corresponding Log₂FC, in blue for downregulated, red for upregulated, black for non-significantly changed in Oat^rhg tissues. (C) Tissue-specific and overlapping enriched pathways in RPE/Cho and liver proteomics, categorized by biological processes and cellular components. Retina was not included as the number of significantly altered proteins was insufficient for enrichment analysis. Endoplasmic reticulum (ER); Mitochondria inner membrane (IM); Major Histocompatibility Complex I (MHC I). ∣log₂fold change∣ ≥ 1, p-value <0.05.

We next compared GO-enriched biological processes and cellular components between RPE/Cho and liver (Fig. 8C). Retina was not included as the number of significantly altered proteins was not sufficient for GO analysis. At the level of biological processes, both tissues showed enrichment of pathways related to lipid metabolism. In RPE/Cho, additional enriched processes were related to transport, carbohydrate metabolism, and the respiratory chain, highlighting broader alterations in RPE/Cho from OAT deficiency. In the cellular component category, both tissues showed overlap in ER-associated proteins. RPE/Cho further showed enrichment of cytoskeletal and mitochondrial components, whereas in the liver, nucleosome- and chromosome-associated proteins were dominant. These differences likely reflect inherent tissue programs, with the liver retaining regenerative capacity, while the post-mitotic RPE/Cho has limited ability to adapt. Overall, these findings indicate that despite minimal overlap in individual proteins and pathways, OAT deficiency drives distinct tissue-specific proteomic changes.

3.9. Cross-tissue metabolomic analysis reveals early shared metabolic alterations in OAT deficiency

To identify common OAT-dependent metabolite changes across liver, RPE/Cho, and retina, we compared significantly altered metabolites in the three tissues using a Venn diagram (Fig. 9A). Only two metabolites were shared among all tissues, and four were shared between RPE/Cho and retina (Fig. 9A and B). Ornithine was consistently increased, whereas N(6)-methyl-lysine decreased in all tissues, linking ornithine accumulation to altered lysine and methylation pathways. This was further supported by reduced levels of 4PY, a nicotinamide-derived metabolite in both RPE/Cho and retina, as well as shared alterations in aminoadipic acid, which was increased in RPE/Cho but decreased in retina. Retina displayed the largest number of altered metabolites (n = 13), compared to 7 in RPE/Cho and liver, suggesting greater metabolic vulnerability to OAT deficiency. KEGG pathway analysis revealed a common disruption of amino-acid metabolism, including arginine biosynthesis, arginine and proline metabolism, and glutathione metabolism, in all three tissues (Fig. 9C). Together, these findings indicate that OAT loss, in addition to altering ornithine-related pathways, leads to early disruptions in one-carbon and redox pathways, with retinal metabolism being the most affected.

Fig. 9. Cross-tissue metabolomic overview of DE proteins in Oat^rhg compared to controls.

(A) Unique and overlapping significantly changed metabolites across tissues in Oat^rhg and (B) showing the overlapping metabolites and their Log₂FC, in blue for downregulated and red for upregulated. (C) Comparison of top enriched metabolic pathways from KEGG analysis in liver, retina, and RPE/Cho. N1-Methyl-4-pyridone-3-carboxamide (4PY). Fold change ≥1.3, p-value <0.05.

4. Discussion

In this study, we have identified early metabolic and proteomic signatures of liver and ocular tissues in OAT deficiency. Ornithine accumulation and reduced N(6)-methyl-lysine are common metabolic features across tissues. The liver metabolome reflects increased ornithine disposal through the urea cycle and impaired SAM-dependent pathways, accompanied by proteomic changes in detoxification enzymes and histone proteins. The retinal metabolome highlights reprogrammed amino acid metabolism to support glutamate homeostasis, with few proteomic changes in cytoskeletal and synaptic proteins. In contrast, RPE/Cho shows the most extensive proteomic remodeling, including downregulation of mitochondrial proteins and alterations in cytoskeletal and ECM-associated proteins. These alterations are accompanied by reduced metabolites linked to mitochondrial function and antioxidant capacity, including creatine, lysine, carnosine, and myo-inositol, indicating early and selective vulnerability of the RPE/Cho.

Beyond ornithine accumulation, a common feature of OAT deficiency in both liver and ocular tissues is the disruption of SAM-dependent metabolism, as indicated by reduced N(6)-methyl-lysine in all tissues. Excess ornithine can be redirected into polyamine synthesis, an important SAM-consuming pathway that uses decarboxylated S-adenosylmethionine (dcSAM) and may limit SAM availability for methylation reactions (Pirini et al., 2025; Schibalski et al., 2024). While SAM-related metabolites and proteins were consistently affected, their downstream effects were tissue-specific. In the liver, our data suggest metabolic compensation by increasing betaine to support SAM regeneration and upregulation of the histone demethylase KDM3B (Bottiglieri, 2002; Barak et al., 2003). These changes in histone methylation patterns likely alter chromatin structure, leading to transcriptional reprogramming, including increased expression of histone H2B proteins (Lim et al., 1993; Li et al., 2017). Moreover, the reduced transsulfuration flux from SAM may impact glutathione levels and sulfate-dependent detoxification (Bottiglieri, 2002).

In contrast to the liver, the eye metabolome and proteome showed limited evidence of metabolic compensation. Reduced levels of 4PY, a methylated nicotinamide byproduct, in both retina and RPE/Cho indicate decreased SAM-mediated methylation. In addition, retinal cystathionine and xanthosine were diminished, further reflecting impaired glutathione synthesis and purine metabolism, both essential for photoreceptor function and survival (Hanna et al., 2022; Plana-Bonamaisó et al., 2020; Álvarez-Barrios et al., 2021; Tokarz et al., 2013). SAM homeostasis depends on de novo synthesis from methionine and recycling from S-adenosyl-L-homocysteine (SAH). In the RPE/Cho, simultaneous downregulation of MAT2B (involved in SAM synthesis) and SAHH2 (SAH hydrolase) suggests a disruption of the methionine cycle. This deficit is corroborated by reduced levels of the SAM-dependent metabolites, including creatinine, 4PY, and N(6)-methyl-lysine. These findings highlight that ocular tissues have a limited capacity to activate alternative pathways to compensate for these disruptions.

The liver detoxifies ammonia through the urea cycle by converting ornithine to citrulline, which is then regenerated from arginine to sustain the cycle. Spatially, urea cycle enzymes are enriched in periportal hepatocytes, whereas OAT, together with cytochrome P450 enzymes involved in detoxification, are restricted to perivenous hepatocytes (Halpern et al., 2017a; Kuo et al., 1991; Jungermann and Katz, 1982). Our proteomic data reveal changes in detoxification proteins that map predominantly to this perivenous zone. The accumulated ornithine and citrulline in the Oat^rhg liver (Fig. 3B) suggest that periportal activity remains intact, possibly shunting excess ornithine into the urea cycle. Although the liver supplies key metabolites necessary for retinal health (e.g., vitamin A, taurine, and unsaturated fatty acids) (Fernández-Sánchez et al., 2011; Mantopoulos et al., 2011; Lawson et al., 2016), it remains to be determined whether such localized disruptions might compromise the availability of factors essential for photoreceptor and RPE survival in GA.

Changes in the retinal metabolome reflect activation of alternative pathways to preserve glutamate. Glutamate functions as the principal excitatory neurotransmitter for phototransduction and a central metabolic intermediate, supporting amino acid metabolism, redox balance, and mitochondrial function (Rego et al., 2000; Bringmann et al., 2013; Satrústegui et al., 2007). To maintain glutamate pools, retina has a high cytosolic reducing power to protect glutamate oxidation through the malate-aspartate shuttle (Du et al., 2013). Glutamate can also be regenerated through transamination of α-KG with other amino acids such as alanine, BCAAs, and lysine. Strikingly, BCAAs, lysine, and its derivative, aminoadipic acid, are reduced in Oat^rhg retinas. BCAAs transamination via branched chain aminotransferases (BCAT1/2) can convert α-KG into glutamate (LaNoue et al., 2001). Inhibition of BCAA transamination and transport in ex vivo rat retina reduced de novo glutamate synthesis by 30% (LaNoue et al., 2001; Aldosari et al., 2024). Beyond their metabolic role, BCAAs are major structural components of visual proteins, comprising ~24% of human rhodopsin (83/354 residues; UniProt P08100) and long-wave cone opsin (85/354 residues; UniProt P04000), with rhodopsin constituting >90% of rod outer segment membrane proteins (Skiba et al., 2023). Thus, the reduction of BCAAs in OAT-deficient retina may impair retinal metabolism and compromise the structural integrity of photoreceptor outer segments.

RPE/Cho shows the most substantial proteomic dysregulation, specifically in mitochondrial metabolism, cytoskeletal, and ECM-related proteins (Fig. 5F). These altered protein expressions may compromise cell bioenergetics, polarity, transport, and barrier integrity, processes closely linked to RPE dysfunction in retinal disease (Ramms et al., 2013; Baek et al., 2017; Wang et al., 2025; Eamegdool et al., 2020; Piskova et al., 2023; Beatty et al., 2000). Focal superonasal RPE lesions have been described in older (12-month-old) Oat^rhg mice (Wilder et al., 2025). Although our analyses were performed on bulk RPE/Cho and at a much earlier disease stage, it will be important in future studies to assess region-specific changes in this mouse model. The mitochondrial proteins involved in oxidative phosphorylation (e.g., NDUBB, NDUS8, COX2), transport (TOM22, MPC1), and fusion (MFN1) are all downregulated, suggesting early mitochondrial dysfunction in Oat^rhg RPE/Cho. Additionally, reduced creatinine and downregulation of mitochondrial creatine kinase (KCRS) further support mitochondrial dysfunction in RPE. Creatinine is a degradation product of creatine, a crucial molecule to support stable ATP supply in energetically demanding tissues via creatine kinase (Wyss and Kaddurah-Daouk, 2000; Acosta et al., 2005). Elevated ornithine is known to inhibit AGAT, the rate-limiting enzyme for creatine synthesis, which may explain reduced creatine availability (Valayannopoulos et al., 2009). These findings align with mitochondrial abnormalities reported in GA patients (McCulloch et al., 1978; Arshinoff et al., 1979; Wilson et al., 1991) and require further studies of mitochondrial respiration and bioenergetics in ocular tissues in GA.

RPE requires robust mitochondria to oxidize nutrients derived from the daily phagocytosis of photoreceptor outer segments and to produce nutrients that support the neural retina (Yam et al., 2019). Mitochondrial dysfunction in RPE is sufficient to induce RPE dedifferentiation, ECM remodeling, and photoreceptor degeneration (Zhang et al., 2021; Zhao et al., 2011; Brown et al., 2019). Unlike other cells, RPE preferentially utilizes proline as a fuel to support mitochondrial metabolism and amino acid synthesis for both RPE and the neural retina (Yam et al., 2019; Zhu et al., 2023a). We speculate that loss of OAT in RPE/Cho may limit proline synthesis by a dual mechanism: reduced proline-derived ornithine through P5C and inhibition of pyrroline-5-carboxylate synthase (P5CS) due to accumulated ornithine (Hu et al., 2008). However, proline levels remained stable in the Oat^rhg RPE/Cho. This is likely compensated by reduced proline oxidation in part due to impaired mitochondria, while simultaneously decreasing proline utilization for ECM and collagen synthesis. Consistently, several ECM proteins are altered in RPE/Cho proteomics. Interestingly, proline supplementation has been reported to slow or halt the progression of chorioretinal lesions in GA (Hayasaka et al., 1985). Further studies are needed to assess the role of proline in RPE mitochondrial dysfunction and protection in GA.

Lysine plays vital roles in post-translational modification, ECM organization, and carnitine metabolism, linking it to epigenetic regulation, lipid, and energy metabolism (Yamauchi and Sricholpech, 2012; Lloyd and He, 2024; Matthews, 2020a; Maas et al., 2020; Wang and Cole, 2020). Hypolysinemia is prevalent in GA patients and has long been attributed to increased renal clearance of lysine (Takki, 1974; Elpeleg and Korman, 2001). Lysine supplementation, which significantly reduces plasma ornithine, has been proposed as a therapeutic strategy for GA (Zekušićc et al., 2018; Aktaşoğlu et al., 2026). Strikingly, we found that OAT deficiency reduces N(6)-methyl-lysine, alters aminoadipic acid, and enhances lysine catabolism in RPE/Cho. Consistently, we previously reported that key enzymes in lysine catabolism (AASS, ALDH7A1, and AADAT) are enriched in human RPE (Puja et al., 2025). This heightened catabolism may increase both lysine influx and downstream product synthesis, potentially contributing to hypolysinemia. Elevated aminoadipic acid is implicated in diabetes, obesity, and renal disease, and correlates with oxidative stress and metabolic disruption (Brown and Kretzschmar, 1998; Estaras et al., 2020; Lee et al., 2019; Sell et al., 2007). Exposure to aminoadipic acid impairs mitochondrial metabolism, antioxidant responses, and cellular transport in human enterocytes (Díaz-Velasco et al., 2020). Lysine is required for collagen cross-linking in the ECM and serves as a precursor for the synthesis of carnitine, a key regulator of lipid metabolism (Maas et al., 2020; Matthews, 2020b; Melegh et al., 1993). Given that defects in ECM remodeling and lipid metabolism are central to retinal degenerative diseases, including inherited retinal degeneration and AMD, dysregulated lysine metabolism in ocular tissues may have important pathogenic consequences (Landowski and Bowes Rickman, 2022; Hussain and Lee, 2025; Coral et al., 2008; Yang et al., 2025; Engel et al., 2022; Liew et al., 2020). Further studies are therefore needed to assess the role of lysine dysregulation in ocular tissues.

In addition, the decreased levels of carnosine and myo-inositol further support impaired RPE metabolism. Carnosine (β-alanyl-L-histidine) is a mitochondria-protective dipeptide that scavenges reactive oxygen species and limits lipid peroxidation (de Almeida Torres et al., 2023; Guo et al., 2019; Bonfanti et al., 1999). Its decline can potentially affect antioxidant capacity and exacerbate mitochondrial function (Caruso et al., 2023a, 2023b; Aloisi et al., 2013) (Fig. 6B-D). Myo-inositol is a precursor for phosphatidylinositol (PI) synthesis. Strikingly, CDP-diacylglycerol-inositol 3-phosphatidyltransferase (CDIPT), a key enzyme for the synthesis of PI, was downregulated in Oat^rhg RPE/Cho (Table S10). Since phosphoinositides are important for regulating membrane trafficking, cytoskeletal organization, mitochondrial function, and outer segment phagocytosis (De Craene et al., 2017; Hou et al., 2025; Lourdes et al., 2024; Chhetri, 2019; Del Monte et al., 1991; Senju and Lappalainen, 2019; Bhattacharya et al., 2016), their downregulation may contribute to the cytoskeletal and mitochondrial defects in Oat^rhg RPE/Cho.

Together, our data demonstrate that OAT deficiency drives early, common, and tissue-specific changes in the metabolome and proteome that extend beyond ornithine metabolism. These findings support the need for therapeutic strategies in GA that target both systemic metabolism and tissue-specific pathways. Approaches aimed at lowering plasma ornithine, together with ocular tissue-targeted metabolic or gene therapies, may be required to preserve visual function and potentially benefit other affected tissues in GA, including the cornea, lens, muscle, and brain (Balfoort et al., 2024, 2025; Jasani et al., 2018; Valayannopoulos et al., 2009; Sipilä, 1980; Bangal et al., 2012).

While our study provides an early molecular signature of liver and eye in OAT deficiency, it has some limitations. First, Oat heterozygous littermates were used as controls rather than wild-type mice. Although heterozygous carriers are reported to be phenotypically normal in this model (Bisaillon et al., 2014), subtle proteomic or metabolomic changes due to reduced Oat gene dosage at the tissue level cannot be fully excluded. Second, proteins were sampled from a single hepatic region, which may not capture the full cellular and functional tissue heterogeneity, particularly given the spatial organization of urea cycle enzymes and OAT (Kuo et al., 1991; Halpern et al., 2017b; Chen et al., 2023). Third, the RPE and choroid were collected together to better preserve RPE integrity. Future studies are required to dissect their tissue-specific changes by optimized separation techniques, staining, or spatial proteomics. Fourth, although OAT expression and activity were experimentally validated, the mechanistic roles of other altered proteins and their associated biochemical pathways are currently inferred and await direct functional validation. Lastly, metabolite data, including ornithine, were reported as relative abundances. Future studies should include absolute quantitation of metabolite concentrations to better assess the effects of OAT deficiency across tissues.

Abbreviations

GA

gyrate atrophy of the choroid and retina

OAT

ornithine aminotransferase

P5C

pyrroline-5-carboxylate

PLP

Pyridoxal 5′-posphate

RPE/Cho

retinal pigment epithelium/choroid

TCA cycle

tricarboxylic acid cycle

ER

endoplasmic reticulum

ECM

extracellular matrix

GO

Gene Ontology

DE

differentially expressed

BCAA

branched chain amino acid

SAM

S-adenosyl methionine

SAH

S-adenosyl homocysteine

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.exer.2026.111140.

Data availability

The mass spectrometry proteomics raw data have been deposited to the ProteomeXchange Consortium via PRIDE partner repository with the dataset identifier PXD063614 and 10.6019/PXD063614. The metabolomics raw data have been deposited to MassIVE repository with identifier MSV000101103.