Signature of miRNAs derived from the circulating exosomes of patients with amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a progressive, fatal neurodegenerative disorder (NDS) with unclear pathophysiology and few therapeutic options. Mutations in and are the most common in Asian and Caucasian patients with ALS, respectively. Aberrant (microRNAs) miRNAs found in patients with gene-mutated ALS may be involved in the pathogenesis of gene-specific ALS and sporadic ALS (SALS). The aim of this study was to screen for differentially expressed miRNAs from exosomes in patients with ALS and healthy controls (HCs) and to construct a miRNA-based diagnostic model to classify patients and HCs. We compared circulating exosome-derived miRNAs of patients with ALS and HCs using the following two cohorts: a discovery cohort (three patients with -mutated ALS, three patients with -mutated ALS, and three HCs) analyzed by microarray and a validation cohort (16 patients with gene-mutated ALS, 65 patients with SALS, and 61 HCs) confirmed by RT-qPCR. The support vector machine (SVM) model was used to help diagnose ALS using five differentially expressed miRNAs between SALS and HCs. A total of 64 differentially expressed miRNAs in patients with -mutated ALS and 128 differentially expressed miRNAs in patients with -mutated ALS were obtained by microarray compared to HCs. Of these, 11 overlapping dysregulated miRNAs were identified in both groups. Among the 14 top-hit candidate miRNAs validated by RT-qPCR, hsa-miR-34a-3p was specifically downregulated in patients with -mutated ALS, while hsa-miR-1306-3p was downregulated in ALS patients with both and mutations. In addition, hsa-miR-199a-3p and hsa-miR-30b-5p were upregulated significantly in patients with SALS, while hsa-miR-501-3p, hsa-miR-103a-2-5p, and hsa-miR-181d-5p had a trend to be upregulated. The SVM diagnostic model used five miRNAs as features to distinguish ALS from HCs in our cohort with an area under receiver operating characteristic curve (AUC) of 0.80. Our study identified aberrant miRNAs from exosomes of SALS and ALS patients with / mutations and provided additional evidence that aberrant miRNAs were involved in the pathogenesis of ALS regardless of the presence or absence of the gene mutation. The machine learning algorithm had high accuracy in predicting the diagnosis of ALS, shedding light on the foundation for the clinical application of blood tests in the diagnosis of ALS, and revealing the pathological mechanisms of the disease.