Objective: To identify and validate circulating micro RNAs (miRNAs) that mark gene expression changes in articular cartilage early in osteoarthritis (OA) pathophysiology process. Methods: Within the ongoing RAAK study, human preserved OA cartilage and plasma (N = 22 paired samples) was collected for RNA sequencing (respectively mRNA and miRNA). Spearman correlation was determined for 114 cartilage genes consistently and significantly differentially expressed early in osteoarthritis and 384 plasma miRNAs. Subsequently, the minimal number of circulating miRNAs serving to discriminate between progressors and non-progressors was assessed by regression analysis and area under receiver operating curves (AUC) was calculated with progression data and plasma miRNA sequencing from the GARP study (N = 71). Results: We identified strong correlations (ρ ≥ |0.7|) among expression levels of 34 unique plasma miRNAs and 21 genes, including 4 genes that correlated with multiple miRNAs. The strongest correlation was between let-7d-5p and EGFLAM (ρ = −0.75, P = 6.9 × 10⁻⁵). Regression analysis of the 34 miRNAs resulted in a set of 7 miRNAs that, when applied to the GARP study, demonstrated clinically relevant predictive value with AUC > 0.8 for OA progression over 2 years and near-clinical value for progression over 5 years- (AUC = 0.8). Conclusions: We show that plasma miRNAs levels reflect gene expression levels in cartilage and can be exploited to represent ongoing pathophysiological processes in articular cartilage. We advocate that identified signature of 7 plasma miRNAs can contribute to direct further studies toward early biomarkers predictive for progression of osteoarthritis over 2 and 5 years.

中文翻译：

---

循环 MicroRNA 与可能反映 OA 易感性的软骨基因表达高度相关——迈向适用的早期 OA 生物标志物的鉴定

目的：鉴定和验证在骨关节炎 (OA) 病理生理过程早期标记关节软骨基因表达变化的循环微 RNA (miRNA)。方法：在正在进行的 RAAK 研究中，收集人类保存的 OA 软骨和血浆（N = 22 对样本）进行 RNA 测序（分别为 mRNA 和 miRNA）。114 个软骨基因在骨关节炎早期和 384 个血浆 miRNA 中一致且显着差异表达，确定了 Spearman 相关性。随后，通过回归分析评估用于区分进展者和非进展者的循环 miRNA 的最小数量，并使用来自 GARP 研究的进展数据和血浆 miRNA 测序计算接受者操作曲线下面积（AUC）（N= 71)。结果：我们确定了 34 个独特的血浆 miRNA 和 21 个基因的表达水平之间存在强相关性（ρ ≥ |0.7|），其中包括与多个 miRNA 相关的 4 个基因。let-7d-5p 和EGFLAM之间的相关性最强（ρ = -0.75，P = 6.9 × 10 ^-5）。对 34 种 miRNA 的回归分析产生了一组 7 种 miRNA，当应用于 GARP 研究时，这些 miRNA 显示出临床相关的预测价值，2 年内 OA 进展的 AUC > 0.8 和 5 年内进展的接近临床价值（AUC = 0.8)。结论：我们表明血浆 miRNA 水平反映了软骨中的基因表达水平，可用于代表关节软骨中正在进行的病理生理过程。我们主张，鉴定出的 7 种血浆 miRNA 的特征有助于进一步研究预测骨关节炎在 2 年和 5 年内进展的早期生物标志物。