Identification and characterisation of temporal abundance of microRNAs in synovial fluid from an experimental equine model of osteoarthritis

Abstract

Background: MicroRNAs, a class of small noncoding RNAs, serve as post-transcriptional regulators of gene expression and are present in a stable and quantifiable form in biological fluids. MicroRNAs may influence intra-articular responses and the course of disease, but very little is known about their temporal changes in osteoarthritis.

Objectives: To identify miRNAs and characterise the temporal changes in their abundance in SF from horses with experimentally induced osteoarthritis. We hypothesised that the abundance of miRNA would change during disease progression.

Study design: In vivo experiments.

Methods: RNA extracted from synovial fluid obtained sequentially (Day 0, 28 and 70) from nine horses with experimentally induced osteoarthritis was subjected to small RNA sequencing using the Illumina Hiseq 4000 sequencing platform. Differentially abundant miRNAs underwent further validation and mapping of temporal abundance (Day 0, 14, 17, 21, 28, 35, 42, 49, 56, 63 and 70 days after osteoarthritis induction) by microfluidic reverse transcription quantitative real-time PCR. Bioinformatic analyses were performed to predict potential biological associations and target genes of the differentially abundant microRNAs.

Results: Small RNA sequencing revealed 61 differentially abundant microRNAs at an early osteoarthritis stage (Day 28), and subsequent reverse transcription quantitative real-time PCR analysis validated 20 of these. Significant biological functions of the differentially abundant microRNAs were apoptosis, necrosis, cell proliferation and cell invasion. Following validation, four microRNAs (miRNA-199b-3p, miRNA-139-5p, miRNA-1839 and miRNA-151-5p) were detected in more than 50% of the synovial fluid samples and had higher abundance in osteoarthritic than in control joints.

Main limitations: Limited sample size.

Conclusion: This is the first study to determine longitudinal changes in synovial fluid microRNA abundance in an equine model of osteoarthritis. Larger studies are needed in naturally occurring osteoarthritis to interrogate putative changes identified by this study.

Keywords: horse; microRNA; osteoarthritis; post‐transcriptional gene regulation; reverse transcription quantitative polymerase chain reaction; small RNA sequencing; synovial fluid.

FIGURE 1

Overview of the experimental workflow. Synovial fluid (SF) samples were obtained from nine horses with experimentally induced osteoarthritis (OA) in the left middle carpal joint (MCJ) using the equine carpal osteochondral fragment model of OA. The right MCJ served as a control joint for each horse (sham surgery). SF obtained on three sampling days from eight horses underwent small RNA sequencing. MicroRNAs of interest, that is, those that were differentially abundant (DA) between Day 0, 28 and 70 (n = 61 miRNAs), were used for computational analysis identifying biological function and prediction of target genes. The 61 DA miRNAs were validated through reverse transcription quantitative polymerase chain reaction (RT‐qPCR), and the temporal abundance pattern mapped by assessing abundance on Day 0, 14, 17, 21, 28, 35, 42, 49, 56, 63 and 70 days after surgery. The RT‐qPCR analyses were undertaken in two steps. First, the 61 miRNAs of interest were analysed in a subset of samples from four horses, yielding 24 miRNA candidates. Second, based on the first RT‐qPCR a panel of 24 miRNAs were selected and analysed in all synovial samples obtained from all nine horses.

FIGURE 2

Histological images experimental induction of osteoarthritis (OA). Representative images from third carpal bone (haematoxylin and eosin staining) 70 days after surgical induction of OA. (A) Articular cartilage from OA joint and (B) articular cartilage from control joint. In the OA joint complete loss of articular cartilage is present, and fibrillating fibrocartilage has replaced the hyaline cartilage.

FIGURE 3

Core analysis of 61 differentially abundant miRNAs identified in synovial fluid from horses with experimentally induced osteoarthritis. The most substantial biological functions identified were (A) apoptosis and necrosis, both of which were predicted to be decreased, (B) cell proliferation and (C) cell invasion, with the latter two predicted to be increased. Key to the main features in the networks is shown. Image made in Ingenuity Pathway Analysis.

FIGURE 4

Temporal abundance of four microRNAs (miRNA). These were present in >50% of synovial fluid samples obtained from nine horses with experimentally induced osteoarthritis (OA), (A) miRNA‐199b‐3b, (B) miRNA‐151‐5p, (C) miRNA‐139‐5p and (D) miRNA‐1839. Sham operated (arthroscopy alone) joints served as controls. Mean relative abundance ±SEM (error bars are depicted). Image made in Prism (GraphPad).