CSF MicroRNAs Reveal Impairment of Angiogenesis and Autophagy in Parkinson Disease

Abstract

Background and objectives: We assessed longitudinal changes in CSF microRNAs (miRNAs) in patients with moderately severe Parkinson disease.

Methods: We used next-generation whole-genome miRNA sequencing to determine CSF miRNA expression in 75 patients with Parkinson disease after single random ascending doses of nilotinib and longitudinal miRNA expression after daily nilotinib, 150 and 300 mg, vs placebo for 1 year.

Results: Significant changes in the expression of miRNAs that control genes and pathways that regulate angiogenesis, autophagy, and the blood-brain-barrier components, primarily collagen, were observed over 1 year, suggesting impairment of these pathways in Parkinson progression in these patients. Different miRNAs that indicate activation of genes associated with autophagy flux and clearance and angiogenesis were significantly altered in the nilotinib, 300 mg vs 150 mg, or placebo group, and these changes correlated with clinical outcomes. No changes were observed in miRNAs after a single dose of nilotinib vs placebo.

Discussion: This study suggests vascular and autophagy defects in Parkinson progression. Nilotinib, 300 mg, reverses these effects via alteration of miRNA expression, suggesting epigenomic changes that may underlie long-term disease-modifying effects.

Trial registration information: Clinical trial registration number: NCT02954978.

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology.

Figures

Figure 1. Cross-sectional Analysis of miRNA After a Single Dose of Nilotinib

(A) Schematic of the clinical trial design. This schematic was created using BioRender.com. (B) Principal component analysis of miRNA expression in the single-dose study for treatment vs untreated effects with placebo- (red) and nilotinib- (blue) treated populations and (C) placebo (red), 150 mg (blue), 200 mg (green), 300 mg (purple), and 400 mg (orange) nilotinib-treated populations. Principal component analysis classified 63 participants into groups using 2,559 miRNAs. PC1 and PC2 are depicted.

Figure 2. Longitudinal MicroRNA Expression

(A) Placebo, 150 mg nilotinib, and 300 mg nilotinib volcano plot of longitudinal miRNA expression in patients with Parkinson disease who underwent placebo treatment. Expression data are plotted as -log10 (p-value) (y-axis) vs log (fold-change) (x-axis). Light colors indicate p < 0.05 for the Wald test, and dark colors indicate that miRNA was also p < 0.05, according to mixed-effects analysis of variance (ME-ANOVA) with the Kruskal-Wallis test. Blue indicates a decrease in expression, and red indicates an increase in expression. The number of miRNAs that meet both the Wald test and ME-ANOVA with the Kruskal-Wallis test (p < 0.05) is shown in the top corners. (B) Heatmap of fold-regulation (FR = FC-1) for miRNAs (p < 0.05) that were in common between the Wald test and Kruskal-Wallis test. ME-ANOVA *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. miRNA that displays an effect of treatment (expression directionality for nilotinib is opposite to placebo) is outlined by dotted line and shown to the right of the heatmap. Tyrosine kinase, autophagy, and angiogenesis predicted gene targets were determined via Ingenuity Pathway Analysis software (Qiagen) with confidence filtering to include only highly predicted gene targets. Gene targets with additional experimental confirmation are indicated by ‡. Dashed lines indicate decreased miRNA expression leading to disinhibition of the gene targets, whereas bold line inhibition indicates increased inhibition of the gene targets.

Figure 3. Pathway and GO Functional Analysis Demonstrate Angiogenesis and BBB Breakdown Associated With Placebo and Inflammation and Autophagy Associated With 300 mg Nilotinib

(A) KEGG (light blue), Reactome (pink), and WikiPathway (light purple) enrichment analysis for placebo differentially expressed miRNAs (DEMs). Enriched pathways were determined using the g:Profiler online tool with Benjamini-Hochberg multiple testing for significance (p < 0.05). (B) Gene ontology (GO) molecular function (green), biological process (orange), and cellular component (purple) enrichment analysis for placebo DEMs. (C) KEGG (light blue), Reactome (pink), and WikiPathway (light purple) enrichment analysis for placebo DEMs. Enriched pathways were determined using the g:Profiler online tool with Benjamini-Hochberg multiple testing for significance (p < 0.05). (D) GO molecular function (green), biological process (orange), and cellular component (purple) enrichment analysis for placebo DEMs. The bubble plot x-axis represents the –log10(Padj), and the size of the bubble corresponds to how many input genes intersect with the associated pathway or GO term.

Figure 4. Correlating Longitudinal miRNA Expression to UPDRS II + III

(A) The sum of Unified Parkinson's Disease Rating Score (UPDRS) II (motor) and III (activities of daily living) and at 0, 12, and 27 months ±SD. The difference in scores is shown below in the table. (B) Heatmap of the correlation between 12-month and longitudinal changes in UPDRS II-III and longitudinal changes in miRNAs were computed using Spearman rank correlations. Placebo group miRNAs that highly correlate with UPDRS II + III (r > 0.5, p < 0.05) are outlined by a dotted line and are shown to the right. Twenty-seven–month correlation when placebo was transitioned to nilotinib, 150 or 300 mg, for the second year (right). (C) KEGG, Reactome, and WikiPathway enrichment analysis of gene regulation by miRNAs that highly correlate with UPDRS II + III in placebo.

Figure 5. Summary of Mechanism for miRNA in Parkinson Disease and Response to DDR1 Inhibition

(A) Discoidin domain receptor1 (DDR1) is overactivated, and several miRNAs, including hsa-miR-30b-5p, hsa-miR-7162-5p, and hsa-miR-4252, regulate several key Parkinson disease-associated pathologies, including blood-brain barrier (BBB) degradation, neuroinflammation, and impaired early autophagy. These pathologies may then promote DDR1 activity, creating a negative feedback loop. When DDR1 is inhibited, via nilotinib, 300 mg, we find that the miRNA differentially expressed miRNAs (DEMs) in placebo are reversed. In addition, we find that miRNAs, including hsa-miR-15a-5p, hsa-miR-5195, and hsa-miR182-5p, positively affect BBB maintenance, transport of glucose across the BBB, and bioenergetics, and restore autophagic flux. This schematic was created using BioRender.com.