We performed an RNA-Seq analysis of tears from 16 Parkinson’s Disease (PD) patients and 16 healthy, age-matched healthy controls (HC) to identify genes and pathways that may serve as biomarkers of PD.
We previously demonstrated that PD tears have significantly higher levels of oligomeric a-synuclein than HC tears. Analyses of hundreds of subjects demonstrated this biomarker predicted PD diagnosis, irrespective of disease duration (AUROC=0.76). To improve this model, we performed RNA-Seq on PD and HC tears. Pilot studies of brain and tears revealed ~74% of genes expressed in aged cortex were expressed in tears.
Schirmer’s Tear Strips were frozen at -80°C. RNA was extracted and evaluated for quantity and integrity (RIN). RNA-Seq libraries were prepared with the NEBNext Ultra II RNA Library Prep Kit and sequenced on an Illumina NovaSeq sequencer. FASTQ files were aligned with Tophat and transcripts were counted and compared with Cufflinks. Pathway analysis was performed using the Qiagen Ingenuity Pathway Analysis (IPA) software and unsupervised, hierarchical clustering of significantly affected pathways was tested with RStudio.
RNA-Seq analysis showed more genes were downregulated in PD tears (N=898) vs. upregulated (N=238). IPA analysis of downregulated genes revealed DNA repair pathways were significantly affected, and reflected the downregulation of 60 genes related to DNA excision repair, recombination, and damage in PD tears. These 60 genes included HNRNPC, TARDBP, USP47, EZH2, FANCC. Unsupervised hierarchical clustering of these 60 genes’ transcript levels (FPKMs) separated PD and HC with 83.9% accuracy (sensitivity 93.3% and specificity 75%).