Fig. 1

Single-cell underpinnings of electropathophysiological activity and possible therapeutic targets. a Depiction of the experimental approach. Left-to-right: surgical retrieval of fresh brain cortical samples for tissue processing, single-cell isolation, single-cell RNA sequencing, and preoperative EEG recordings. After clustering of the integrated single-cell RNA sequencing data, cell types were annotated according to the expression of known marker genes and visualized using t-stochastic neighborhood embedding (tSNE) map right; bar plot distribution by percentage of the identified cell types between cohorts. Here, the recoding of electroencephalographic data used to extract beta and narrow gamma-band oscillations as PD electropathophysiologic hallmarks is also depicted. Partly created with BioRender.com. b Differential gene expression analysis between PDb and non-PD (n = 5) depicted as enrichment network maps per cell type in microglia (left), astrocytes (middle), and oligodendrocytes precursor cells (OPCs; right), highlighting metabolic and inflammatory pathways abnormally regulated in PD. The size of the circle corresponds to the count of genes in each term, color-coded by normalized enrichment score (NES), and showing only statistically significant results (p FDR-adjusted < 0.05). c Compared to non-PD (n = 38; light blue), PDb (n = 9; dark green) showed robustly mirrored the same increased beta power (0.28 ± 0.12, T = 2.68, p = 0.005) and reduced narrow gamma power (0.62 ± 0.15; T = 2.34, p = 0.012) as found in larger PD samples (n = 91; light green; Beta PD [mean ± sd] 0.22 ± 0.15 vs non-PD 0.15 ± 0.11; T = 2.77, p = 0.003; narrow gamma PD 0.7 ± 0.21 vs non-PD 0.77 ± 0.19; T = 1.8, p = 0.037). Group differences in phase-amplitude coupling showed an increase in PD (0.06 ± 0.03; T = 1.81, p = 0.036) in comparison to non-PD (0.05 ± 0.03), and marginal significance in PDb (0.05 ± 0.02; T = 1, p = 0.05). d Gene-drug interaction analysis using enrichR showed strong drug-gene interactions between 5 genes, with the greatest group differences (PDb vs non-PD), and NVP-BEP800 (q = 0.000002), an ATP-competitive inhibitor of Heat-shock protein 90 (HSP90), which could be used for the development of new treatment strategies for PD. Bonferroni correction within each cell type was used to correct the p-value. EEG electroencephalography, OPCs oligodendrocytes precursor cells, PD Parkinson’s disease, NES normalized enrichment score, PDb PD subjects with brain biopsies