Background/Purpose: MicroRNAs have been shown to contribute to osteoarthritis (OA) pathophysiology, yet little is known about the circulating miRNome in OA. The circulating miRNome (e.g. all microRNAs in blood plasma) may include promising biomarkers of disease given that microRNAs are relatively stable, easy to detect, and easy to quantify. Since there are currently no validated biomarkers for detecting early stages of OA, profiling the circulating miRNome in different stages of OA holds potential for biomarker discovery. Next generation sequencing is an advantageous approach for profiling microRNAs because it offers the sensitivity and specificity to detect novel and low abundance microRNAs that are potentially unique to disease stages. Therefore, we hypothesize that sequencing will identify unique signatures of circulating microRNAs in symptomatic patients with early radiographic knee OA as compared to late radiographic knee OA. 

Methods: Early knee OA patients were defined based on radiographic Kellgren-Lawrence grades 0 and 1 (N=41), and late knee OA patients defined based on Kellgren-Lawrence grades 3 and 4 (N=50). All patients were symptomatic and showed clinical features of knee OA. Of note, patients with Kellgren-Lawrence grade 2 were excluded in an effort to clearly demarcate early from late stage OA. Plasma from each patient (N=91) was subjected to microRNA library preparation and sequencing on the Illumina NextSeq550 platform. Sequencing reads were aligned and counts were generated for both known microRNAs (documented in miRBase v22.1) and novel microRNAs (predicted using bioinformatic tools). Demographic, anthropometric, and clinical data collected for all patients were considered as covariates in statistical analyses. Further statistical, bioinformatics, and computational biology approaches were taken to refine and interpret the final list of microRNAs and their predicted gene targets. 

Results: Sequencing data were first explored in an unbiased manner using principal component (PC) analysis. This revealed clear separation of samples according to OA stage, with late OA samples forming a distinct cluster from early OA samples (PC1 = 58.2%). Differential expression analysis identified 215 microRNAs at FDR < 0.01 in early OA versus late OA. Refining this list for microRNAs that were consistently increased or decreased across ≥ 85% of samples in the early OA group as compared to their median expression in the late OA group, we found 97 microRNAs. At a threshold of  ≥ 95%, 7 microRNAs were identified, one of which was consistently found to be increased in 100% of early OA samples. Exploring novel microRNAs, 4 were found in  ≥ 50% of early OA samples. Interestingly, gene target predictions showed several common targets between the 97 shortlisted microRNAs and the 4 novel microRNAs, including SMAD2 and associated TGF-beta signaling pathway. 

Conclusion: Sequencing the circulating miRNome identified a unique signature of 11 microRNAs, which included 4 novel microRNAs, in patients with early knee OA. Future studies should be directed towards understanding the role, mechanisms, and utility of these microRNAs as OA biomarkers. 

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/mirnome-sequencing-identifies-a-unique-profile-of-circulating-micrornas-in-early-knee-osteoarthritis/