Serum non-coding RNAs as biomarkers for osteoarthritis progression after ACL injury

Abstract

Objective: The aim of this study was to examine serum non-coding RNAs as potential biomarkers for cartilage damage associated with anterior cruciate ligament (ACL) injury.

Methods: Serum was obtained from 80 patients 1 year after surgery for ACL injury and 60 normal donors without overt skeletal injury. Total serum RNA was isolated, small non-coding RNAs profiled by TaqMan array MicroRNA (miRNA) analysis and individual small RNA assays performed by quantitative TaqMan RT-PCR (qPCR). Semi-quantitative magnetic resonance imaging (MRI) analysis was performed using Whole Organ Magnetic Resonance Knee Score (WORMS) scoring for analysis of cartilage damage.

Results: Initial TaqMan array miRNA profiling showed an increased serum concentration of a small nucleolar RNA (snoRNA), U48, in five patients with cartilage damage compared with that in five patients without cartilage damage and six normal donors. Independent qPCR analysis of snoRNAs in serum from all patients and normal donors showed a strong association between the serum level of another snoRNA, U38, and cartilage damage in ACL injury patients and together with snoRNA, U48, clear distinction between ACL injury patients and normal donors.

Conclusion: SnoRNAs U38 and U48 are significantly elevated in the serum of patients developing cartilage damage at 1 year after ACL injury. Serum levels of U38 have the potential to facilitate early diagnosis of patients with cartilage damage after ACL injury. This study suggests serum non-coding RNAs may serve as novel noninvasive biomarkers for the detection and assessment of cartilage damage after ACL injury.

Figure 1

Comparison of miRNA concentration in serum from ACL injury patients with and without cartilage damage and serum from normal donors A, Comparison of miRNA levels in serum from ACL injury patients developing cartilage damage (OA, WORMS score 13 to 35) and patients without evidence of cartilage damage (Control, WORMS score 0). B, comparison of miRNA levels in serum from ACL injury patients with cartilage damage and normal donors. miRNAs were analyzed by ABI TaqMan array after RNA isolation from 0.4 ml serum. Each point represents the average and standard deviation of the cycle threshold (Ct) of a miRNA from 5 patients or 6 normal donors. Points below the normal distribution, such as that shown in red, have higher levels in serum from patients developing cartilage damage. The ncRNA shown in red is snoRNA U48.

Figure 2

The association between serum snoRNA U48 and U38 levels and cartilage pathology. The copy number of snoRNAs in serum from patients after ACL injury or normal donors was determined by RT-PCR after reverse transcription. Cartilage damage was determined by detailed MRI using WORMS scoring. A, Comparison of cartilage damage with patients WORMS score (r_s=0.13, p=0.27 for U48 and r_s =0.38, p=0.001 for U38). B, the data were binned according to the level of cartilage damage measured by WORMS score for ACL injury patients (0, 1–3, 4–8 and >8). Copy number in agematched normal donors (normal) is included separately. n=number of donors. For snoRNA U38, serum levels between age-matched control or patients with WORMS scores of 0 or 1–3 and patients with WORMS score > 8 all had Wilcoxon p < 0.05, For snoRNA U48, all age-matched control comparisons to ACL groups had p<0.05. None of the ACL group to group differences were significant (p>0.24).

Figure 3

Receive Operator Characteristic Plot. The data shown in Fig. 2 were used to draw a ROC plot comparing; U38 serum levels in ACL injury patients with no or very mild cartilage damage (WORMS score ≤3) and patients with greater cartilage damage (WORMS score ≥4).

Figure 4

Age association of serum concentration of U48. The serum concentration of U48 (copy #) from ACL injury patients (open circles) or normal donors (black diamonds) is plotted against the age of the donor.