Multi-omic analysis reveals lipid dysregulation associated with mitochondrial dysfunction in Parkinson’s disease brain.

Hällqvist J, Toomey CE, Pinto R, Baldwin T, Doykov I, Wernick A, Shahrani MA, Evans JR, Lachica J, Pope S, Heales S, Eaton S, Mills K, Gandhi S, and Heywood WE. (2025) Multi-omic analysis reveals lipid dysregulation associated with mitochondrial dysfunction in Parkinson’s disease brain. Nature Communications

Parkinson’s disease (PD) is a progressive neurodegenerative disease that is characterized by tremors, muscle stiffness, slowed movements, and difficulty walking. PD is the most common neurodegenerative movement disorder with 1-2% of people over 65 years of age being diagnosed with the disorder, totalling over 6.1 million people worldwide. Dysregulation in lipid metabolism pathways has been associated with the etiology of PD. Although there is also research into the influence of lipids on alpha-synuclein aggregation, literature is lacking on how lipid profiles differ throughout the disease progression. Hällqvist et al were interested in investigating the lipid profile in the brains of people with PD and determining what the differences are between the stages of the disease and across brain regions. To accomplish this, they performed targeted lipidomics on eight brain regions in post-mortem controls individuals, patients with mid-stage PD, and patients with late-stage PD.

To determine the baseline levels of lipids that would be expected in healthy, aged brains, a group of 107 control samples were analysed. Spatial differences were found when looking at lipid distribution across the brain and they noted that ceramide levels changed the least between brain regions. In contrast, lyso-phosphatidylethanolamine and n-hexosylceramide demonstrated the most variation at 82% and 52%, respectively. Hierarchical clustering was used to examine any correlations between lipid levels across the brain and determine whether different regions have similar lipid profiles or not. In these comparisons they found that there were three clusters: 1) frontal cortex, caudate, and parahippocampus; 2) cerebellum, cingulate cortex, and putamen; and 3) parietal and temporal cortices. The species that drove this clustering were a greater abundance of phospholipids and sphingolipids in the first group and elevated levels of ganglioside in the second group. The third group demonstrated lower levels of most lipids except lyso-phospholipids and ceramide. These findings indicate that there are regional differences throughout the brain, but each smaller section is not particularly distinct from those near it.

After learning the lipid distribution to expect in a healthy aged brain, the authors wanted to evaluate what lipid changes occur in the brains of patients with PD when compared to a healthy brain. A principal component analysis (PCA) model of PD and control samples across brain regions was developed. This demonstrated that gangliosides and long-chain polyunsaturated phospholipids were elevated in PD brains. These also formed groups that were distinguishable between the brain regions. All eight regions assessed were enriched in different lipid classes and, interestingly, the cerebral cortex had greater levels of ceramide, phospholipids, and ethanolamine plasmalogens while the inner brain regions had greater levels of ganglioside and lyso-n-hexosylsphingosine. To evaluate the region-specific differences between people with PD and healthy controls, hierarchical clusters were used. Of these, two clusters were found with the first consisting of the parahippocampus, cingulate, temporal frontal, and parietal cortices and the second containing  the putamen, caudate, and cerebellum. This suggests that there are different changes in PD between the brain regions and that the changes are likely related to their structure and function. The first cluster contains regions of the cerebral cortex and the limbic system and here gangliosides and several hexosylceramides were elevated in PD while phospholipids were lower. In the second cluster phospholipid choline and lyso-phospholipid cholines were higher in PD samples than the control while glycosphingolipids were lower. These differences in the clusters could suggest neuroanatomical distinctions in the PD brain and different pathological profiles between regions since the clusters have unique roles. The cortical cluster (first cluster) contains structures involved in cognitive processing, memory, and sensorimotor integration and is usually affected in late stages of PD, while the second cluster contains regions that are more closely tied to PD pathology and are affected in the early stages.

To further confirm whether the lipid changes they had seen aligned with the progression of the disease, the samples were separated according to Braak staging. The mid-stage of PD was Braak 3-4 and the late stage was Braak 5-6. Three brain regions (cerebellum, frontal cortex, and putamen) were evaluated for the comparison between the PD samples and the control samples. In Braak’s staging of PD progression, the putamen is the most affected area during mid-stage PD because of its role in the basal ganglia and its involvement in dopaminergic signalling. Interestingly, the frontal cortex is typically not affected at mid-stage, but demonstrates distinct pathological changes in late-stage PD, which aligns with the cognitive and executive function impairment that occurs in this stage of the disease. The authors found significant differences in lipid levels between mid- and late-stage PD, largely driven by an increase in lyso-phospholipids, ceramide, ganglioside, and poly-unsaturated diacylated phospholipids in the late-stage samples. The putamen demonstrated significant changes in both stages with the mid-stage having reduced sphingolipids while shorter chain C18 ceramides were elevated. Late-stage had elevated levels of a variety of phospholipids and sphingolipids.

Hällqvist et al wanted to determine how the lipid profile differs in the brains of people with PD and how brain region and disease progression influences this. They confirmed levels of lipids across eight brain regions in healthy control samples to act as a baseline for the expected lipid distribution across the brain and they found three clusters in the brain with distinct lipid profiles. They then wanted to compare the lipidome of a PD diseased brain and a healthy brain and found that gangliosides and polyunsaturated phospholipids are increased in PD brains. The authors also demonstrated that the lipid changes in different brain regions aligned with the Braak staging of the disease. In mid-stage PD, the putamen is the most effected region and showed the most alterations to sphingolipid and ceramide levels in this stage, while in late-stage PD the frontal cortex is affected more and they found elevated levels of many phospholipids and sphingolipids in this region. Together these results support the role of lipid changes in Parkinson’s disease and indicate that these alterations are specific to the stage of the disease. Further work in this area could provide new pathways for treatment options in this disease.