A rapidly progressive multiple system atrophy-cerebellar variant model presenting marked glial reactions with inflammation and spreading of α-synuclein oligomers and phosphorylated α-synuclein aggregates

Abstract

Multiple system atrophy (MSA) is a severe α-synucleinopathy facilitated by glial reactions; the cerebellar variant (MSA-C) preferentially involves olivopontocerebellar fibres with conspicuous demyelination. A lack of aggressive models that preferentially involve olivopontocerebellar tracts in adulthood has hindered our understanding of the mechanisms of demyelination and neuroaxonal loss, and thus the development of effective treatments for MSA. We therefore aimed to develop a rapidly progressive mouse model that recaptures MSA-C pathology. We crossed Plp1-tTA and tetO-SNCA*A53T mice to generate Plp1-tTA::tetO-SNCA*A53T bi-transgenic mice, in which human A53T α-synuclein—a mutant protein with enhanced aggregability—was specifically produced in the oligodendrocytes of adult mice using Tet-Off regulation. These bi-transgenic mice expressed mutant α-synuclein from 8 weeks of age, when doxycycline was removed from the diet. All bi-transgenic mice presented rapidly progressive motor deterioration, with wide-based ataxic gait around 22 weeks of age and death around 30 weeks of age. They also had prominent demyelination in the brainstem/cerebellum. Double immunostaining demonstrated that myelin basic protein was markedly decreased in areas in which SM132, an axonal marker, was relatively preserved. Demyelinating lesions exhibited marked ionised calcium-binding adaptor molecule 1-, arginase-1-, and toll-like receptor 2-positive microglial reactivity and glial fibrillary acidic protein-positive astrocytic reactivity. Microarray analysis revealed a strong inflammatory response and cytokine/chemokine production in bi-transgenic mice. Neuronal nuclei-positive neuronal loss and patchy microtubule-associated protein 2-positive dendritic loss became prominent at 30 weeks of age. However, a perceived decrease in tyrosine hydroxylase-positive neurons in the substantia nigra pars compacta in bi-transgenic mice compared with wild-type mice was not significant, even at 30 weeks of age. Wild-type, Plp1-tTA, and tetO-SNCA*A53T mice developed neither motor deficits nor demyelination. In bi-transgenic mice, double immunostaining revealed human α-synuclein accumulation in neurite outgrowth inhibitor A (Nogo-A)-positive oligodendrocytes beginning at 9 weeks of age; its expression was further increased at 10 to 12 weeks, and these increased levels were maintained at 12, 24, and 30 weeks. In an α-synuclein-proximity ligation assay, α-synuclein oligomers first appeared in brainstem oligodendrocytes as early as 9 weeks of age; they then spread to astrocytes, neuropil, and neurons at 12 and 16 weeks of age. α-Synuclein oligomers in the brainstem neuropil were most abundant at 16 weeks of age and decreased thereafter; however, those in Purkinje cells successively increased until 30 weeks of age. Double immunostaining revealed the presence of phosphorylated α-synuclein in Nogo-A-positive oligodendrocytes in the brainstem/cerebellum as early as 9 weeks of age. In quantitative assessments, phosphorylated α-synuclein gradually and successively accumulated at 12, 24, and 30 weeks in bi-transgenic mice. By contrast, no phosphorylated α-synuclein was detected in wild-type, tetO-SNCA*A53T, or Plp1-tTA mice at any age examined. Pronounced demyelination and tubulin polymerisation, promoting protein-positive oligodendrocytic loss, was closely associated with phosphorylated α-synuclein aggregates at 24 and 30 weeks of age. Early inhibition of mutant α-synuclein expression by doxycycline diet at 23 weeks led to fully recovered demyelination; inhibition at 27 weeks led to persistent demyelination with glial reactions, despite resolving phosphorylated α-synuclein aggregates. In conclusion, our bi-transgenic mice exhibited progressively increasing demyelination and neuroaxonal loss in the brainstem/cerebellum, with rapidly progressive motor deterioration in adulthood. These mice showed marked microglial and astrocytic reactions with inflammation that was closely associated with phosphorylated α-synuclein aggregates. These features closely mimic human MSA-C pathology. Notably, our model is the first to suggest that α-synuclein oligomers may spread from oligodendrocytes to neurons in transgenic mice with human α-synuclein expression in oligodendrocytes. This model of MSA is therefore particularly useful for elucidating the in vivo mechanisms of α-synuclein spreading from glia to neurons, and for developing therapies that target glial reactions and/or α-synuclein oligomer spreading and aggregate formation in MSA.