Serum small non-coding RNA define molecular subtypes in amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease with variable site of onset, disease progression rates and survival times. Early-stage ALS characteristics are shared with other conditions, posing diagnostic challenges and resulting in diagnosis delays. We investigated tRNA-derived small RNAs (tsRNAs) and microRNAs (miRNAs) which are stable and abundantly expressed small non-coding RNAs (sncRNAs) as potential diagnostic serum biomarkers, comparing them to healthy controls and ALS mimics, and gained pathophysiological insights from dysregulated sncRNAs. We analyzed small RNA-seq data from 158 patients with ALS, 60 healthy controls and 39 patients with neurological conditions that mimic ALS to identify differentially expressed sncRNAs. A classifier was built to evaluate their diagnostic potential, followed by hierarchical clustering to identify ALS molecular subtypes. Finally, we performed gene ontology and pathway analysis to identify pathways disrupted within subtypes. We identified several dysregulated tsRNAs and miRNAs and assessed their diagnostic potential using an extreme gradient boosting (XGBoost) classifier. Our models achieved an accuracy of 87.16% and 82.23% in classifying patients with ALS from healthy controls and ALS mimics, respectively. We identified four sncRNA expression-based ALS molecular subtypes with one C9orf72 enriched cluster. Further analysis of identified differentially expressed sncRNAs showed their involvement in neuronal pathways. Our study identified potential sncRNA-based diagnostic serum biomarkers and associated molecular subtypes which can be further studied to match clinical parameters and develop subtype specific biomarkers and therapeutic strategies for ALS.肌萎缩侧索硬化症（Amyotrophic lateral sclerosis, ALS）是一种致命性神经退行性疾病，其发病部位、疾病进展速率与生存时长均存在显著个体差异。ALS早期的临床表现与其他多种疾病存在重叠，这给临床诊断带来了极大挑战，进而导致诊断延迟。本研究将稳定且高丰度表达的小非编码RNA（small non-coding RNAs, sncRNAs）家族中的转运RNA衍生小RNA（tRNA-derived small RNAs, tsRNAs）与微RNA（microRNAs, miRNAs）作为潜在血清诊断生物标志物，以健康对照者及ALS模拟症患者为对照展开研究，并从失调的小非编码RNA中解析疾病的病理生理学机制。我们共分析了158例ALS患者、60例健康对照者及39例ALS模拟症神经系统疾病患者的小RNA测序（small RNA-seq）数据，以筛选差异表达的小非编码RNA。随后构建分类器评估其诊断潜力，并通过层级聚类鉴定ALS的分子亚型。最后，通过基因本体论与通路分析，明确各亚型中发生紊乱的信号通路。本研究筛选出多种失调的tsRNAs与miRNAs，并采用极端梯度提升（extreme gradient boosting, XGBoost）分类器评估其诊断效能。结果显示，本研究模型在区分ALS患者与健康对照者、区分ALS患者与ALS模拟症患者时，分别达到了87.16%与82.23%的分类准确率。我们鉴定出4种基于小非编码RNA表达谱的ALS分子亚型，其中包含1个富集C9orf72的聚类簇。对筛选得到的差异表达小非编码RNA的进一步分析显示，它们参与了神经元相关通路的调控。本研究鉴定出了潜在的血清小非编码RNA类诊断生物标志物及相关分子亚型，后续可结合临床参数开展深入研究，以开发针对ALS的亚型特异性生物标志物与治疗策略。