Altered exosomal miRNA profiles in patients with paraneoplastic cerebellar degeneration

Abstract

Objective: Patients with ovarian cancer (OC) may develop anti-Yo-associated paraneoplastic cerebellar degeneration (PCD)-a cerebellar ataxia associated with tumor-induced autoimmunity against CDR2 and CDR2L proteins. Dysregulation of circulating exosomal microRNAs (miRNAs) occur in OC. Here, we investigated whether PCD is associated with changes in the exosomal miRNA profiles of OC patients.

Methods: Serum exosomes were isolated from patients with OC (n = 15), patients with OC and anti-Yo-associated PCD (n = 14) and healthy controls (HC, n = 15). Small RNA sequencing was used to identify differentially expressed miRNAs. Receiver operating characteristic curves were used to evaluate biomarker sensitivity and specificity, and miRNA target prediction analysis was employed to elucidate gene targets.

Results: OC patients with PCD exhibited a distinct exosomal miRNA expression profile. We detected 103 differentially expressed exosomal miRNAs in PCD patients compared to OC patients without PCD and 139 differentially expressed exosomal miRNAs compared to controls. Particularly miR-486-5p, miR-4732-5p, miR-98-5p and miR-21-5p exhibited notable sensitivity and specificity for discriminating PCD patients from both OC patients without PCD and healthy controls. miRNA target prediction showed that several of the differentially expressed miRNAs in PCD patients targeted the CDR2 and CDR2L genes.

Interpretation: Our results demonstrate that OC patients with anti-Yo-associated PCD exhibit a distinct exosomal miRNA profile compared to OC patients without PCD. Several of the differentially expressed exosomal miRNAs in PCD patients showed diagnostic potential and may hold relevance for understanding the pathogenesis of PCD.

Figure 1

Characterization of exosomal RNA expression. Exosomes were isolated from sera from ovarian cancer (OC) patients, ovarian cancer patients with PCD, and healthy controls (HC). (A) The total read distribution of RNA types detected by small RNA sequencing. (B) The proportion of miRNA reads mapping to each chromosome. (C) The miRNAs with the most reads in each sample. (D) Principal component analysis plot of the miRNA expression. Each point represents a sample and are color‐coded according to sample group. The arrows indicate how strongly each miRNA influences the principal components.

Figure 2

Differentially expressed exosomal miRNAs. (A) Volcano plots of the significantly (FDR <0.05) up‐ (red) and downregulated (blue) miRNAs for each group comparison (panels). (B) Venn diagram of differentially expressed miRNAs between all group comparisons. The intersects represents differentially expressed miRNAs common between group comparisons. (C) Bar plot showing the log2 fold changes (log2FC) of the 14 miRNAs differentially expressed both in OC vs HC and PCD vs HC. HC, healthy controls; OC, ovarian cancer; PCD, paraneoplastic cerebellar degeneration.

Figure 3

Exosomal miRNAs as biomarkers for ovarian cancer and PCD. ROC curve analysis was performed to assess the diagnostic performance of exosomal miRNAs in discriminating PCD patients from OC patients and healthy controls (A), and OC patients from PCD patients and healthy controls (B). The left panels show the ROC curves where the true positive rate (sensitivity) is plotted against the false positive rate (1—specificity). The right panels show the expression levels as determined by small RNA sequencing and normalized using log2 counts per million (CPM). The horizontal lines indicate the median ± SD. AUC, area under the curve; HC, healthy controls; OC, ovarian cancer; PCD, paraneoplastic cerebellar degeneration.

Figure 4

Functional enrichment analysis of the genes targeted by the differentially expressed miRNAs. Enrichment analysis was performed using DIANA‐miRPath on the differentially expressed miRNAs between PCD and OC (n = 103), PCD and healthy controls (n = 139), and OC and healthy controls (n = 16) (panels). The y‐axis represents the enriched KEGG gene sets, the x‐axis represents the false discovery rate (FDR) on the negative log10 scale, and the transparency of the bars represents the number of differentially expressed miRNAs targeting genes in the enriched gene sets. HC, healthy controls; OC, ovarian cancer; PCD, paraneoplastic cerebellar degeneration.