Novel blood lncRNA biomarkers associated with clinical severity and specific cognitive dimensions in Alzheimer's disease

Abstract

Background Differential expression of long non-coding RNAs (lncRNAs) in brain, serum, and blood show strong potential to distinguish Alzheimer's disease (AD) from healthy controls. Objective To explore whether lncRNA signatures delineate AD pathology and map to distinct, multidimensional cognitive domains, enhancing specificity in assessing AD severity and progression. Methods We profiled 29,603 lncRNAs transcripts in blood samples from 15 AD patients and 15 healthy controls, alongside comprehensive neuropsychological assessments. Generalized Linear Models and Predictive Power Score analyses, with statistical prioritization, identified lncRNAs associated to AD neuropsychological architecture. Results Several lncRNAs share strongly associated with cognitive performance and AD severity, mapping to genes involved in key AD-related molecular processes, including synaptic and neurotransmitter regulation (e.g., EPHB1, CHRNA4, TEAD1), protein homeostasis and Aβ pathology (e.g., FBXL2, FAM221A, APP), mitochondrial function and cellular stress (e.g., VDAC3, PPT2-EGFL8), neuroinflammation and immune regulation (e.g., TEAD1, EMX2OS, LY6E-DT), epigenetic and transcriptional control (e.g., PRDM2, DLEU1, FIRRE), neuronal excitability (e.g., KCNJ14), and neuroprotection and synaptic plasticity (e.g., SIL1). Novel associations included ferroptosis, DNA stability, microtubule dynamics, and dendritic orientation (e.g., BTB3, DICER1, GNG7, IBA57, NEAT1, POT1, SRD5A3). Conclusions We identify candidate lncRNA signatures that may serve as potential biomarkers and enhance our understanding of the molecular basis of the cognitive architecture in AD, opening new avenues for biomarker identification and targeted therapeutic strategies development. Validation in larger, diverse cohorts is essential to confirm their mechanistic contributions to AD.