SOD1 Aggregates Injected into Brain Fluid Trigger ALS in Mice

Researchers have shown that misfolded aggregates of the protein superoxide dismutase-1 (SOD1), injected into the cerebrospinal fluid of mice, can travel from that fluid into the central nervous system and trigger a fatal, ALS-like disease far earlier than it would otherwise appear. The finding, published in Acta Neuropathologica Communications, demonstrates for the first time that the cerebrospinal fluid can serve as a transmission route for SOD1 aggregates, which behave in a prion-like manner by templating the misfolding of their normally shaped counterparts.

The team used transgenic mice expressing a mutant human SOD1 variant, G85R, which in their laboratory lives to about 405 days on average without showing signs of disease until late in life. That long symptom-free window made the animals well suited for detecting whether an injected seed could accelerate aggregation.

Two structurally distinct strains of human SOD1 aggregates, labeled A and B, had previously been identified in transgenic mice. Earlier work showed that both strains could transmit templated aggregation and induce early-onset fatal ALS-like disease when injected directly into the spinal cord or a peripheral nerve of susceptible mice. Peripheral administration into the abdominal cavity or skeletal muscle, by contrast, did not induce disease, suggesting that SOD1 aggregates do not spread through the bloodstream or across neuromuscular junctions. What remained unknown was whether the cerebrospinal fluid, which bathes the brain and spinal cord, could carry the aggregates back into nervous tissue.

To test this, the researchers inoculated 100-day-old, still-healthy G85R transgenic mice into the cisterna magna, a fluid-filled space at the base of the skull where cerebrospinal fluid is readily accessible. A first attempt using 1 microliter of a strain A homogenate, containing about 1.6 micrograms per milliliter of SOD1 aggregates, produced no significant difference in survival compared with uninoculated mice. The team reasoned that the aggregates would be diluted in the cerebrospinal fluid, so they developed a new seed preparation designed to concentrate very high molecular weight aggregate species, isolating proteinaceous components larger than one million kilodaltons using density-cushion centrifugation.

With the concentrated seeds, the results were striking. Mice receiving 5 microliters of strain A seeds, roughly 9 micrograms per milliliter of aggregates, showed the first signs of disease 87 days after inoculation, while mice receiving strain B seeds, at about 13 micrograms per milliliter, showed signs at 92 days. Both groups reached the terminal disease stage approximately 180 days earlier than mice that received a control preparation made from non-transgenic mouse spinal cords. No difference in disease course emerged between the two strains, and control-inoculated mice did not differ in survival or phenotype from uninoculated animals, whose reference survival curve was drawn from 287 historical mice.

Binary epitope mapping, a technique using eight antibodies raised against overlapping segments of the SOD1 sequence that bind only to disordered, misfolded protein, confirmed that the injected seeds had templated aggregation in the recipients. Strain A-inoculated mice expressed a strain A pattern, and two of three strain B-inoculated mice expressed a typical strain B pattern, with the N-terminus folded into the fibril core and the remainder protruding in a disordered manner. Because strain B does not form spontaneously in the G85R mouse line, its appearance indicated genuine prion-like transmission of the seed structure. Quantitative mapping along the neuroaxis showed the highest aggregate levels in the cervical spinal cord for most mice, with the brainstem highest in one animal from each group, matching the predominantly forelimb symptoms observed at the end stage.

Immunohistochemistry revealed cytoplasmic condensed SOD1 aggregates in neurons throughout the ventral horn of the spinal cord and brainstem in the inoculated mice, with strain A animals also showing perinuclear and nuclear staining. Age-matched control-inoculated mice, examined at about 250 days, showed only an evenly distributed granular cytoplasmic staining, and only developed condensed aggregates when they reached terminal disease at roughly 400 days. The researchers also noted an absence of inflammation in the examined central nervous system tissue, with no granulocytes in the tissue proper and no large clusters of round cells or activated microglia, indicating that the pathology was not driven by an immune reaction at the injection site.

The authors emphasize that the doses used in the mice far exceed the concentrations of misfolded SOD1 measured in human cerebrospinal fluid. Accounting for an estimated total mouse cerebrospinal fluid volume of about 37 microliters, a 5 microliter inoculation is diluted roughly 7.4-fold, yielding a final concentration of approximately 1.35 micrograms per milliliter of hSOD1 in the fluid, compared with about 0.12 micrograms per milliliter of total hSOD1 reported in human cerebrospinal fluid. Misfolded SOD1 has been detected in the cerebrospinal fluid of ALS patients and in ALS transgenic rats, but its concentration there is very low, roughly one 1300th of total SOD1, which has so far precluded structural analysis of the human material. The team also observed that a fivefold smaller dose produced no effect on survival within the observed timeframes, and that a homogenate preparation with lower aggregate content produced slower symptom onset and longer survival, consistent with a dilution effect in the cerebrospinal fluid.

Several limitations frame the findings. The work was performed entirely in a transgenic mouse model overexpressing mutant human SOD1, and the mice were inoculated at 100 days of age with laboratory-prepared seeds. The study design could not distinguish between directional propagation of the aggregates from the cerebrospinal fluid into tissue and multifocal seeding, a question the authors leave to future research. Batch variability also surfaced unexpectedly: among three separately prepared strain A seed batches, two were more potent than the third despite nearly identical SOD1 protein content, a difference the authors suspect may arise from the homogenization and sonication steps fragmenting high molecular weight aggregates to different extents. Six mice whose lifespans resembled uninoculated animals, suggesting failed inoculations, were excluded from the statistical calculations, and four mice were excluded altogether for surgical or unrelated health reasons.

Nevertheless, the conclusion is clear: SOD1 aggregates can enter the central nervous system from the cerebrospinal fluid, induce premature ALS-like disease, spread along the neuroaxis in a prion-like manner, and form cytoplasmic inclusions in neurons. The highest aggregate burden appeared in the brainstem and cervical spinal cord, regions adjacent to the inoculation site, while the cerebrum and cerebellum, though equally close, contained far sparser aggregates, suggesting that certain regions are especially susceptible to SOD1 aggregation. The authors propose that insufficient control of SOD1 structure and aggregation may underlie the vulnerability of motor areas, noting that autophagy factors are lower in the motor neuron pools of the ventral horn. If the cerebrospinal fluid route of spread applies beyond the mouse model, SOD1 in the fluid could become a therapeutic target, and mapping the transmission pathways of pathological protein aggregates may open new avenues for preventing the spread of ALS and other neurodegenerative diseases.

Subject of Research: Transmission of SOD1 protein aggregates from cerebrospinal fluid into the central nervous system, inducing ALS-like disease in transgenic mice

Article Title: Intracisternal injection of two human SOD1 aggregate strains induces templated aggregation and ALS in transgenic mouse model

Article References: Sigfridsson, I., Marklund, M., Andersen, P. M., Zetterström, P., Marklund, S. L., & Brännström, T. (2026). Intracisternal injection of two human SOD1 aggregate strains induces templated aggregation and ALS in transgenic mouse model. Acta Neuropathologica Communications, 14(1), Article 225. https://doi.org/10.1186/s40478-026-02473-x

Image Credits: AI Generated

DOI: 10.1186/s40478-026-02473-x

Keywords: ALS, SOD1, cerebrospinal fluid, prion-like spread, templated aggregation, transgenic mice, neurodegeneration, protein aggregates, motor neurons, cisterna magna, G85R, misfolded protein