A new study has revealed that the retina and brain handle the same disease-linked prion protein in strikingly different ways, offering fresh clues about why some neurological proteins form destructive amyloid plaques in one region of the central nervous system but remain in a less organized, non-amyloid state in another. The research, published in Acta Neuropathologica, examined tissues from people affected by Gerstmann–Sträussler–Scheinker disease, or GSS, a rare inherited prion disorder caused by mutations in the PRNP gene. The investigators focused on the F198S mutation, which replaces the amino acid phenylalanine with serine at position 198 of the prion protein. Their findings suggest that the retina is not simply a transparent “window” into the brain. Instead, it may provide a biologically distinct environment in which abnormal proteins aggregate, are processed and acquire disease-associated properties in different ways.
GSS is a dominantly inherited neurodegenerative disease in which abnormal prion protein accumulates over years or decades. The F198S form is associated with progressive behavioral changes, cognitive impairment, cerebellar ataxia, parkinsonism and pyramidal motor signs. In the brain, the disease produces characteristic PrP deposits, including diffuse accumulations and plaques with dense amyloid cores. These plaques are made largely from fragments of the prion protein that have adopted highly ordered, beta-sheet-rich structures. Such structures bind dyes including thioflavin S, a standard laboratory marker of amyloid. The same mutation can also be accompanied by tau neurofibrillary tangles, creating a pathology that combines features of prion protein amyloidosis and tau-related neurodegeneration. Until now, however, little was known about what the F198S protein does in the retina.
The research team studied postmortem brain tissue from eight F198S mutation carriers and retinal tissue from seven of them. The individuals belonged to the same extended pedigree, an important feature that reduced some of the genetic and clinical variability that can complicate studies of rare inherited diseases. The researchers also examined a case of GSS caused by a different mutation, A117V, as well as retinal and brain samples from people with Alzheimer’s disease, dementia with Lewy bodies and sporadic Creutzfeldt–Jakob disease. Using conventional histology, immunohistochemistry, fluorescence microscopy, biochemical fractionation and real-time quaking-induced conversion, or RT-QuIC, the scientists compared the location, structure, solubility, protease resistance and seeding activity of abnormal PrP in the two tissues.
The anatomical pattern in the retina was unusually precise. In every GSS retina examined, abnormal PrP deposits were found in the outer plexiform layer, or OPL, and not in the other retinal layers. The OPL is a synaptically dense region where photoreceptors communicate with bipolar and horizontal cells. Under the microscope, the deposits appeared as small bead-like structures rather than the large plaques seen in the brain. In two extensively analyzed cases, the deposits averaged about 7.75 micrometers in diameter. They were located close to ribbon synapses, specialized structures that allow photoreceptors to continuously transmit visual information, and they overlapped with the dendrites of rod bipolar cells. The deposits did not colocalize with microglia, and the tissue did not show evidence of widespread retinal inflammation, gliosis or loss of retinal layers.
Although the retinal beads reacted strongly with nine antibodies recognizing different regions across the full length of PrP, they behaved differently from classical amyloid. They did not fluoresce with thioflavin S, nor did they fluoresce with the luminescent conjugated oligothiophenes pFTAA and HS-84. These specialized probes can bind particular arrangements of protein aggregates and have been used to distinguish molecular forms of amyloid in diseases involving PrP, tau and amyloid-beta. The negative signal indicates that the retinal deposits lack the ordered amyloid architecture found in the cores of cerebral PrP plaques. The result was similar in the A117V GSS retina, where deposits were also restricted to the OPL and were not detected by the amyloid-sensitive dyes. In contrast, brain plaques from F198S cases displayed the expected thioflavin S fluorescence, especially in the cerebellar cortex, where PrP pathology was particularly abundant.
The biochemical analysis provided an explanation for this morphological divide. In the brain, the pathological material included a prominent eight-kilodalton internal PrP fragment, commonly called PrP internal fragment or PrPIF. This fragment spans approximately residues 80 to 150 of the protein and corresponds closely to the sequence forming the core of F198S amyloid filaments identified in earlier cryo-electron microscopy studies. In brain samples, the fragment accumulated not only as a monomer but also as covalently linked multimers ranging from approximately 16 to 35 kilodaltons. These multimers are thought to assemble into the amyloid structures that give GSS plaques their characteristic staining and physical properties.
The retina contained a different molecular mixture. Full-length PrP was present, including highly glycosylated forms that were more prominent in F198S tissue than in control retina. The retinal protein was partly detergent-insoluble and formed large aggregates, but its distribution across sucrose density gradients differed from that of brain-derived PrP. A distinctive 25-kilodalton glycosylated form was detected in the retina but not the brain. The retinal samples also contained C-terminal fragments known as C1 and C2. Yet the constitutive eight-kilodalton PrPIF that was readily detectable in the brain was absent from untreated retinal homogenates. When the investigators exposed retinal material to proteinase K, an enzyme commonly used to probe the resistance of misfolded PrP, an eight-kilodalton fragment appeared. Under highly denaturing conditions, the retinal protein could be dissociated into an eight-kilodalton species, suggesting that the fragment was generated experimentally from larger retinal PrP molecules rather than naturally accumulated in the tissue.
The difference was also reflected in the ability of the samples to seed the conversion of normal PrP into abnormal forms. RT-QuIC uses repeated shaking and incubation to amplify tiny quantities of disease-associated PrP. The reaction contains a recombinant PrP substrate and thioflavin T, whose fluorescence increases as newly formed aggregates appear. Because the molecular compatibility between a seed and a recombinant substrate can vary, the study tested several substrates. The clearest responses occurred with bank vole PrP carrying the V109M sequence. Brain samples from F198S patients remained positive at dilutions as high as one million-fold, whereas retinal samples produced detectable reactions down to a hundred-thousand-fold but not at one-million-fold dilution. The estimated seeding activity was approximately two logarithmic units lower in the retina than in the brain. This does not mean that retinal PrP was biologically inert; rather, it indicates that the retinal aggregates were less efficient at initiating the particular conversion measured by this assay.
The findings point toward a tissue-specific mechanism controlling prion protein processing. The F198S mutation increases the structural instability of PrP and makes it more prone to misfolding, but the mutation alone does not determine the final shape of every aggregate. Cellular composition, glycosylation, membrane organization, protease activity, synaptic architecture and local clearance systems may all influence the path taken by the protein. In the brain, an endogenous proteolytic process may generate the internal fragment that subsequently polymerizes into amyloid fibrils. The retina may lack the relevant protease, express it at a different level or place PrP in a molecular environment that prevents the same cleavage. The OPL is rich in synaptic membranes and specialized signaling machinery, factors that could stabilize full-length PrP or redirect its aggregation into non-amyloid assemblies.
The study does not establish whether the retinal deposits impair vision, because the patients had not undergone systematic testing with electroretinography or optical coherence tomography during life. Nor can it determine whether the retinal changes arise early in disease or mainly represent late-stage pathology. The biochemical experiments were performed on only two individuals, and the overall sample size was necessarily limited by the rarity of F198S GSS and the difficulty of obtaining well-preserved eyes after death. Nevertheless, the consistency of the OPL pattern across multiple affected individuals strengthens the conclusion that the retina is a distinct site of PrP pathology. Future research will need to determine whether the retinal deposits are inside synaptic compartments or outside cells, whether they alter photoreceptor-to-bipolar-cell signaling and whether their molecular structure can be resolved by cryo-electron microscopy. The authors also propose systematic retinal imaging and electrophysiological studies in living PRNP mutation carriers. If measurable retinal abnormalities emerge before severe neurological decline, the eye could become a valuable, minimally invasive site for monitoring inherited prion disease. For now, the central message is clear: the same mutant protein can follow dramatically different biochemical routes in two connected parts of the nervous system, revealing how local biology may decide whether misfolded PrP remains an unstable aggregate or advances toward amyloid.
Subject of Research: Prion protein aggregation in the brain and retina in inherited Gerstmann–Sträussler–Scheinker disease
Article Title: Distinctive properties of the prion protein in the brain and retina in the amyloidosis associated with the PRNP F198S mutation
Article References: Ghetti, B., Glazier, B. S., Fiorini, M. et al. “Distinctive properties of the prion protein in the brain and retina in the amyloidosis associated with the PRNP F198S mutation.” Acta Neuropathologica 152, Article 25 (2026).
Image Credits: AI Generated
DOI: 10.1007/s00401-026-03060-z
Keywords: Gerstmann–Sträussler–Scheinker disease, prion protein, PRNP F198S mutation, retina, brain, amyloid, PrP internal fragment, RT-QuIC, retinal synapses, neurodegeneration
Tags: Amyloid pathology in prion diseasesAmyloid plaque formation in brain and retinaDifferential prion protein aggregation mechanismsInherited prion disorders and genetic mutationsNeurodegeneration in Gerstmann–Sträussler–Scheinker diseasePrion protein in neurodegenerative diseasesPrion protein processing in central nervous systemPRNP F198S mutation in GSSRole of retina as a prion disease biomarkerStructural differences of prion protein in brain vs retina
