Sequencing identifies a distinct signature of circulating microRNAs in early radiographic knee osteoarthritis

Authors

S. A. Ali, University Health Network University of Toronto
R. Gandhi, University Health Network University of Toronto
P. Potla, University Health Network University of Toronto
S. Keshavarzi, University Health Network University of Toronto
O. Espin-Garcia, University Health Network University of Toronto
K. Shestopaloff, University Health Network University of Toronto
C. Pastrello, University Health Network University of Toronto
D. Bethune-Waddell, University Health Network University of Toronto
S. Lively, University Health Network University of Toronto
A. V. Perruccio, University Health Network University of Toronto
Y. R. Rampersaud, University Health Network University of Toronto
C. Veillette, University Health Network University of Toronto
J. S. Rockel, University Health Network University of Toronto
I. Jurisica, University Health Network University of Toronto
C. T. Appleton, Western UniversityFollow
M. Kapoor, University Health Network University of Toronto

Document Type

Article

Publication Date

11-1-2020

Journal

Osteoarthritis and Cartilage

Volume

28

Issue

11

First Page

1471

Last Page

1481

URL with Digital Object Identifier

10.1016/j.joca.2020.07.003

Abstract

Objective: MicroRNAs act locally and systemically to impact osteoarthritis (OA) pathophysiology, but comprehensive profiling of the circulating miRNome in early vs late stages of OA has yet to be conducted. Sequencing has emerged as the preferred method for microRNA profiling since it offers high sensitivity and specificity. Our objective was to sequence the miRNome in plasma from 91 patients with early [Kellgren–Lawrence (KL) grade 0 or 1 (n = 41)] or late [KL grade 3 or 4 (n = 50)] symptomatic radiographic knee OA to identify unique microRNA signatures in each disease state. Design: MicroRNA libraries were prepared using the QIAseq miRNA Library Kit and sequenced on the Illumina NextSeq 550. Counts were produced for microRNAs captured in miRBase and for novel microRNAs. Statistical, bioinformatics, and computational biology approaches were used to refine and interpret the final list of microRNAs. Results: From 215 differentially expressed microRNAs (FDR < 0.01), 97 microRNAs showed an increase or decrease in expression in ≥85% of samples in the early OA group as compared to the median expression in the late OA group. Increasing this threshold to ≥95%, seven microRNAs were identified: hsa-miR-335-3p, hsa-miR-199a-5p, hsa-miR-671-3p, hsa-miR-1260b, hsa-miR-191-3p, hsa-miR-335-5p, and hsa-miR-543. Four novel microRNAs were present in ≥50% of early OA samples and had 27 predicted gene targets in common with the prioritized set of predicted gene targets from the 97 microRNAs, suggesting common underlying mechanisms. Conclusion: Sequencing of well-characterized patient cohorts produced unbiased profiling of the circulating miRNome and identified a unique panel of 11 microRNAs in early radiographic knee OA.