The patient was a 91 year-old man with a long-standing history of dopamine-responsive Parkinsonism. He presented with dysphagia approximately 14 years prior to death. He then developed an asymmetric resting tremor, reduced arm swing, stooped posture, and a shuffling gait which responded to Carbidopa-Levodopa therapy. He was clinically diagnosed with Parkinson’s disease 13 years prior to his death. His motor symptoms gradually worsened, he suffered repeated falls, and eventually developed orthostatic hypotension.

As part of his participation in the Baltimore Longitudinal Study of Aging (BLSA) and the Johns Hopkins Alzheimer’s Disease Research Center (JHADRC), he underwent repeated neuroimaging and cognitive testing. The last magnetic resonance imaging (MRI) study, approximately 5 years prior to death, showed only moderate age-related parenchymal volume loss. Based on clinical and cognitive evaluations at serial BLSA visits, the participant was judged to be cognitively normal (based on standard diagnostic consensus criteria) through the final visit, which occurred approximately 2 years prior to death. He had no known history of toxin exposure and no history of psychiatric symptoms or anti-psychotic therapy. There was no family history of movement disorders or dementia. As part of the JHADRC protocol, the patient underwent an autopsy restricted to the brain after death.

At autopsy, the gross brain weight was 1450 g. The left hemisphere, brainstem, and cerebellum were sectioned fresh, while the right hemisphere was fixed in formalin for other studies. No gross photographs were obtained. There was no significant cortical atrophy noted, though there was mild ex vacuo hydrocephalus of the lateral ventricle. There was no evidence of atrophy or discoloration of the basal ganglia. Diagnostic sections were taken from fresh tissue and the remaining tissue was frozen and banked. On gross examination, the locus ceruleus and substantia nigra were grossly well pigmented and histologic examination showed an age-appropriate population of pigmented neurons. Hematoxylin and eosin (H&E) staining showed a mildly prominent Bergmann glial cell layer in the cerebellar cortex and neurofibrillary tangles in the hippocampus, entorhinal cortex, and locus ceruleus. There was no overt neuronal loss, atrophy, or glial inclusions on H&E staining of the basal ganglia. Modified Bielschowsky silver staining showed a primary age-related tauopathy (PART) (Braak IV) [1], with no evidence of amyloid deposition by silver staining or amyloid immunostaining (6 F/3D, DAKO, #M0872, 1:100). There were no Lewy bodies present on hematoxylin and eosin staining, and no accumulation of alpha-synuclein in neurons or glia by immunohistochemistry (42, BD Transduction Biosciences, #610,786, 1:100). While there was mild arteriolosclerosis, the was no evidence of ischemic vascular pathology excluding vascular Parkinsonism.

Ubiquitin immunostaining (Ub-1, Chemicon, #MAB1510, 1:2000) showed diffuse glial ubiquitin-positive inclusions which were most prominent in the basal ganglia (Fig. 1a). Areas of higher density of glial inclusions in the basal ganglia were associated with gliosis by GFAP immunostaining (DAKO, #Z0334, 1:500) and an increased density of activated microglia by Iba-1 staining (Wako, #019-19741, 1:800) (Fig. 1b-c). The next most densely affected areas were the cerebral cortex and the Bergmann glial cell layer in the cerebellar cortex (Fig. 1d-e). A control case with PART showed no evidence of ubiquitin-positive glial pathology in the mesial temporal cortex (Fig. 1f). Immunostaining of multiple cases of PART with either no cerebellar pathology or Bergmann gliosis associated with cerebellar infarcts showed no evidence of ubiquitin-positive glial inclusions (Fig. 1g). In the cortex, all layers were affected, though the deeper layers tended to have more dense glial pathology compared to the superficial layers. There was no subpial, subependymal, or perivascular arrangement of the inclusions.

Pathologic Findings in Ubiquitin-Positive Astrogliopathy. Ubiquitin-positive glial inclusions are most prominent in the basal ganglia (a), associated with GFAP-positive gliosis (b) and increased Iba-1-positive activated microglia (c). Ubiquitin-positive glial inclusions also affected cortex (d), and the Bergmann glial layer of the cerebellar cortex (e). Control cases with primary age-related tauopathy do not show ubiquitin-positive glial inclusions in the mesial temporal cortex (f) or associated with Bergmann gliosis over a chronic infarct in the cerebellum (g). Ubiquitin-positive glial inclusions are typically fine linear aggregates in glial processes (h) with scattered glia also showing dense aggregates in the cell body (i). Linear accumulation in the glial processes of the cerebellum is prominent in the Bergmann glial cell layer (j). Scale bars are 50 μm (g) and 20 μm (h-j). Scale bar in g applies to a-g

The glial aggregates were typically fine linear inclusions in glial processes (Fig. 1h), though there were scattered glia with more dense cell body accumulation (Fig. 1i) which were most common in the basal ganglia. The cerebellar glial inclusions were distinctive, occurring solely in the cerebellar Bergmann glial cell layer and showing a roughly radial orientation toward the surface of the cerebellar cortex (Fig. 1j). The ubiquitin-positive glial inclusions spared the white matter, thalamus, and brainstem, including the substantia nigra. No definite pathologic inclusions were identified in neurons, aside from those associated with PART. Phosphorylated tau immunostaining (Fig. 2a) (AT8, Research Diagnostics, Inc. #MN1020B, 1:50) confirmed PART but was negative in the glial aggregates and did not show age-related tau astrogliopathy (ARTAG) [3]. Another commonly used phosphorylated tau antibody (PHF1, Peter Davies lab, 1:200) showed a similar staining pattern (data not shown). p62 immunostaining (3, BD Transduction Laboratories, #610,833, 1:50), performed in multiple brain regions in this case, also highlighted neurofibrillary tangles but was negative in the glial cytoplasmic inclusions (Fig. 2b). FUS (Fig. 2c) (ag2150, ProteinTech, #11570-1-AP, 1:50) and TDP-43 (Fig. 2d) (ProteinTech, #10782-2-AP, 1:2000) were negative in the glial inclusions and showed no evidence of neuronal pathology.

Other Pathologic Findings. Phospho-tau immunostaining (AT8) shows primary age-related tauopathy involving neurons (a), but does not show aggregates in glia. Immunostaining for p62 is negative in the glial aggregates (b). Immunostaining in the hippocampal dentate gyrus shows no evidence of FUS (c) or TDP-43 (d) aggregation. Scale bar is 50 μm (d) and applies to a-d