Induction of retinopathy by fibrillar oxalate assemblies

The study investigates whether oxalate, which accumulates in primary hyperoxaluria, can form non-crystalline supramolecular structures that contribute to retinal pathology. While metabolite crystals (e.g., calcium oxalate) are implicated in disease, direct clinical evidence for toxic metabolite fibrils has been lacking. Primary hyperoxaluria type 1 leads to systemic oxalate accumulation and has been associated with retinal crystalline deposits and visual impairment, though mechanisms remain unclear and reports are variable. The authors observed impaired retinal function in pediatric hyperoxaluria patients without detectable retinal crystals, prompting the hypothesis that oxalate may self-assemble into fibrillar structures that are cytotoxic and pathogenic to the retina. The purpose is to demonstrate oxalate fibril formation, assess their cytotoxicity, evaluate in vivo retinal effects, and determine their presence and immunogenicity in patients.

Prior work shows certain metabolites can self-assemble into toxic fibrillar assemblies, notably phenylalanine in phenylketonuria, with deposits observed post mortem in patients. Metabolite crystals are known to drive inflammation and pathology in several diseases, including calcium oxalate in hyperoxaluria and kidney stone pathogenesis. Retinal crystalline deposits have been described in hyperoxaluria type 1, sometimes leading to severe visual impairment with progressive retinal damage, but retinal findings are rare and variable due to disease rarity. The authors’ previous studies and others extended the amyloid hypothesis to non-proteinaceous metabolite assemblies (e.g., phenylalanine, tryptophan), which can form ordered fibrils and induce cytotoxicity, suggesting a possible unified mechanism in inborn errors of metabolism.

Clinical evaluation: Five children with genetically confirmed primary hyperoxaluria type 1 underwent comprehensive ophthalmic assessment at Rambam Health Care Campus, including best-corrected visual acuity, slit-lamp exam, dilated fundus examination, fundus autofluorescence, and macular SD-OCT (Spectralis). Functional testing included full-field electroretinography (ff-ERG) after 20 minutes dark adaptation using UTAS 3000 per ISCEV standards, and visual evoked potentials (VEP; UTAS-4000). ERG a-wave and b-wave amplitudes and implicit times were measured; VEP P100 and flash parameters assessed. Ethics approvals and consent were obtained.

Preparation of oxalate assemblies: A heating–cooling self-assembly method was used. For calcium oxalate crystals: 10 mg/ml calcium oxalate monohydrate in PBS heated 4 h at 90 °C to supersaturation, cooled overnight at room temperature, centrifuged 5 min at 20,000 × g; crystals collected from supernatant/pellet interface. For fibrils: the same supersaturated solution was immediately centrifuged post-heating, supernatant collected and cooled overnight to form fibrils. Sodium oxalate fibrils were formed by heating 2 mg/ml sodium oxalate in PBS for 4 h at 90 °C then cooling overnight. Unassembled oxalate: 2 mg/ml oxalate in PBS.

Biophysical characterization: Transmission electron microscopy (TEM; JEM-1400Plus, 80 kV) imaged assemblies. Energy-dispersive X-ray spectroscopy (EDS) probed calcium content in crystals vs fibrils.

Generation and validation of anti-oxalate fibril antibodies: Rabbits were immunized with oxalate fibrils; polyclonal IgG purified via protein G; specificity validated by dot blot against fibrils and lack of binding to unassembled oxalate.

Antibody purification from human sera: Sera diluted 1:5 in PBS, applied to protein G column; bound antibodies eluted with 0.1 M glycine (pH 3.0), neutralized with Tris-HCl, buffer-exchanged to PBS, concentration via BCA, purity by SDS-PAGE.

Dot-blot immunoassays: Preformed oxalate fibrils fixed on nitrocellulose using Bio-Dot apparatus; blocked with 5% skim milk; incubated with primary antibodies (1:200; purified from patient or control sera, or rabbit anti-oxalate fibrils) and HRP secondary (1:10,000); developed with Luminata Forte, imaged on Amersham Imager 600. Control antigen: tetanus toxin to confirm antibody titers in controls.

Cell studies: ARPE-19 and HEK-293 cells seeded at 2 × 10^5 cells/ml in 96-well plates. Oxalate assemblies prepared in serum-free DMEM/F12. Cells treated overnight with oxalate fibrils (from calcium or sodium oxalate), calcium oxalate crystals, unassembled oxalate, or alanine controls (~6 mM). Viability assessed by MTT assay (570 nm readout, 650 nm background). Internalization: ARPE-19 grown on poly-L-lysine coverslips, treated with oxalate fibrils, fixed, permeabilized, blocked, stained with rabbit anti-oxalate fibrils (1:100) and Cy3 secondary (1:200); DAPI counterstain; confocal imaging and Z-stacks performed.

Animal model: Adult male Sprague-Dawley rats received intravitreal injections: 10 µl oxalate fibrils in PBS into one eye; contralateral eye received PBS vehicle. Anesthesia and mydriasis per protocol. Indirect ophthalmoscopy confirmed proper injection. ERG recorded at baseline, 7 and 30 days post-injection. For ERG analysis, a- and b-wave maximal amplitudes (Vmax) were calculated using a hyperbolic model; experimental/control eye Vmax ratios computed to control for technical factors. Separate cohort (n=3) received monomeric oxalate as control. Thirty days post-injection, rats were sacrificed and retinas processed for immunohistochemistry.

Immunohistochemistry: Eyes fixed (2% paraformaldehyde, 2.5% glutaraldehyde), cryoprotected in sucrose, embedded in OCT, sectioned at 16 µm. Sections blocked (3% serum, 0.1% Triton X-100 in PBS), incubated with rabbit anti-oxalate fibrils (1:500) overnight, Rhodamine Red-X secondary (1:500), DAPI (1:2000). Confocal imaging (Zeiss LSM 700); fluorescence quantified in Fiji. GFAP staining assessed Müller cell activation (data referenced).

• Clinical observations: In five pediatric primary hyperoxaluria type 1 patients, ophthalmoscopy and SD-OCT showed normal retinal structure without detectable crystals, yet two children exhibited significantly reduced dark-adapted full-field ERG a- and b-wave amplitudes, indicating impaired global retinal function despite absence of crystalline deposits.
• Oxalate fibril formation: Differential centrifugation of supersaturated calcium oxalate solutions yielded two populations—predominantly fibrils vs crystals. TEM revealed elongated nanofibrils. EDS detected calcium in crystals but not in fibrils, indicating calcium-free oxalate fibrils. Sodium oxalate alone also formed fibrils, confirming calcium-independent fibrillization.
• Cytotoxicity: ARPE-19 viability decreased to ~60% with oxalate fibrils from calcium oxalate and ~55% with fibrils from sodium oxalate (both at ~6 mM), similar to calcium oxalate crystals (~60%). Unassembled oxalate (~90% viability) and alanine (~95%) showed minimal toxicity. In HEK-293 cells, oxalate fibrils reduced viability to ~55–60%, crystals to ~75%, while unassembled oxalate (~95%) and alanine (~90%) were not toxic. ARPE-19 cells internalized oxalate fibrils, confirmed by specific antibody staining and Z-stacks.
• Immunodetection in patients: Antibodies purified from sera of all five hyperoxaluria patients recognized preformed oxalate fibrils in dot blot, but not unassembled oxalate. Antibodies purified from five healthy controls did not recognize oxalate fibrils; robust responses to tetanus toxin verified adequate titers in controls. Using a customized anti-oxalate fibril antibody, dot blots detected oxalate fibrils in patient sera but not in healthy controls, indicating systemic presence and immunogenicity of oxalate fibrils in disease.
• In vivo retinal dysfunction: Intravitreal oxalate fibrils in rats (n=10) caused significant reduction of ERG a- and b-wave Vmax in treated eyes vs control eyes, with experimental/control Vmax ratios ~0.65 for both waves at 7 and 30 days (p<0.01), persisting through 30 days. Monomeric oxalate injections (n=3) did not affect ERG up to 30 days, supporting fibril-specific toxicity.
• ERG pattern similarity: Rat ERG after fibril injection showed a marginally electronegative configuration (decreased b/a ratio) at 30 days, resembling the abnormal ERG relationship observed in affected patients.
• Retinal localization: Immunostaining in treated rat retinas showed intense oxalate fibril signal along photoreceptor outer segments at 2 and 30 days, and strong labeling of Müller glial cells at 30 days. GFAP immunoreactivity indicated Müller cell activation. Control eyes lacked staining.

The findings address the discrepancy between impaired retinal function and absence of detectable retinal crystals in pediatric hyperoxaluria by demonstrating that oxalate can self-assemble into calcium-free fibrillar structures that are cytotoxic to retinal cells, impair retinal function in vivo, and are present and immunogenic in patients. The in vitro and in vivo data support a causative role for oxalate fibrils in oxalate-associated retinopathy, extending the paradigm of metabolite amyloid-like assemblies to a clinical context. The rat model reproduced ERG abnormalities akin to those seen in patients, and tissue staining localized fibrils to photoreceptor outer segments and Müller glia, consistent with functional deficits and glial activation. This work suggests that beyond crystal deposition, fibrillar metabolite assemblies may represent key pathogenic agents in inborn errors of metabolism and provides a rationale for targeting metabolite fibrillization as a therapeutic strategy.

This study provides, to the authors’ knowledge, the first direct clinical indication and in vivo evidence that metabolite fibrillar assemblies—specifically calcium-independent oxalate fibrils—form under defined conditions, are immunogenic and present systemically in hyperoxaluria patients, and cause retinal cytotoxicity and functional impairment mimicking patient phenotypes. The work shifts the mechanistic understanding of oxalate-associated disease beyond crystals to include fibrillar assemblies and establishes a rat model for testing interventions. Future research should elucidate the ultrastructural properties of oxalate fibrils, dissect molecular mechanisms of retinal toxicity, evaluate systemic formation and distribution dynamics, and develop therapeutics that prevent or disrupt oxalate fibrillization to mitigate retinopathy and potentially other organ involvement.

• Small clinical cohort (five pediatric patients) with variable disease severity; only two showed ERG abnormalities, limiting generalizability.
• The in vivo model uses acute intravitreal administration of preformed fibrils rather than reproducing chronic systemic oxalate accumulation; translational relevance to natural disease kinetics requires caution.
• Structural characterization of fibrils is limited to TEM and EDS; detailed ultrastructure, polymorphism, and molecular packing remain to be resolved.
• Observation window was up to 30 days in rats; longer-term retinal outcomes and reversibility were not assessed.
• Mechanistic pathways of cytotoxicity and cell-type specific effects beyond Müller glia activation were not fully delineated.