Long Non-Coding RNA PVT1 and Its Target miRNA-146a as Potential Prognostic Biomarkers in Rheumatoid Arthritis Patients

Objective: Long non-coding RNAs (lncRNAs) and their target microRNAs were documented in multiple studies to have a significant role in different joint disorders such as rheumatoid arthritis (RA) and osteoarthritis (OA). The current work aimed to determine the potential role of lnc-PVT1 and miR-146a as promising biomarkers to distinguish between RA, OA patients, and healthy individuals. Methods: The expression levels of lnc-PVT1 and its target miR-146a in the serum were measured for three different groups, including patients with RA (40), OA patients (40), and healthy controls (HCs) (40). Participating individuals were subjected to a full history investigation and clinical examination. Blood samples were tested for ESR, RF, CBC, as well as liver and renal functions. Serum was used to detect the relative expression levels of lnc-PVT1 and miR-146a and we correlated the levels with RA and OA activity and severity signs. Results: Lnc-PVT1 expression level was greater among patients with RA compared to that of OA patients, with a fold change median of 2.62 and 0.22, respectively (p = 0.001). The miR-146a fold change was significantly demonstrated between the RA, OA, and HCs groups. There was no correlation between both biomarkers with the disease activity scales (DAS28) of RA, the Knee injury Osteoarthritis Outcome Score (KOOS), or any sign of detection of the disease severity of OA. Conclusions: lnc-PVT1 and miR-146a could be considered as promising biomarkers for the diagnosis of RA and OA and may have an important role as therapeutic targets in the future.


Introduction
Rheumatoid arthritis (RA) is a common chronic autoimmune disease of the joints. RA is characterized by cartilage destruction, chronic inflammation of joints (mainly those of the hands and feet), positive rheumatoid factor (serum), as well as abnormal synovial hyperplasia, which subsequently causes joint deformities and dysfunction [1][2][3]. Currently, the exact cause of RA remains unknown but it is usually associated with other systemic

Materials and Methods
The local ethics committee of Fayoum University Hospital approved the current study protocol (Code: R183). Accordingly, all procedures followed the standards of the Declaration of Helsinki.

Study Design and Patients
This study included 80 patients (40 with RA and a similar number of OA patients) and 40 healthy volunteers. Patients were selected from the Department of Rheumatology at Fayoum University Hospital, Egypt. The control group consisted of 40 volunteers who had no family history of RA or any autoimmune disorder. Before being enrolled, all patients were asked to fill out an informed consent form. Any patient with an autoimmune disease, cancer, a chronic infectious disease, or a new infection within a month of enrollment was excluded.

Rheumatoid Arthritis Patients
The study included 40 patients with RA according to the 2010 American College of Rheumatology (ACR)/European League against Rheumatism (EULAR) criteria [28]. All patients have undergone a complete history investigation, clinical examination, assessment of the tender and swollen joint count (SJC/TJC), as well as assessment of the disease activity score (DAS-28) [29]. Any assessment in the form of a questionnaire was given in an Arabic version. The health assessment questionnaire was calculated (ranging from 0 to 2) [30]. Laboratory investigations included liver function tests, renal function tests, rheumatoid factor, as well as erythrocyte sedimentation rate (ESR). Regarding radiographic evaluation, antero-posterior plain radiographs of both hands and wrists were done to determine the joint space narrowing (JSN), erosion, and osteopenia as illustrated in the schematic outline ( Figure S1a).

Osteoarthritis Group
OA patients were diagnosed using the ACR criteria [31] supported by radiographic evidence. Patients excluded from the study were those that presented with any sign of secondary OA, muscle strain involving the lower extremities, neurological disorders, inflammatory arthritis, and ligament sprain. Radiographic imaging was used to assess the degree of involvement, including the presence of JSN, the formation of a subchondral cyst, and any marginal spur formation. A blinded (from the study) radiologist evaluated the results of imaging using the Kellgren and Lawrence (KL) grading, which has high intra-rater reliability [32]. This grading ranges from a scale of (0) to (4), where (0) refers to no radiological indications of OA, (1) indicates doubtful osteophyte, (2) indicates definite osteophyte, (3) indicates the presence of moderate JSN, and (4) indicates severely narrowed joint spacing. The Knee injury and Osteoarthritis Outcome Score (KOOS) Arabic version was used [33]. For detecting disease severity, 6 MWT [34], second Chair Stand Test, and Stair Climb test were performed [35] as illustrated in the schematic outline ( Figure S1b).

Data Collection and Laboratory Investigation
At registration, the following data was collected from patients and controls: age, gender, clinical presentation, and BMI. The blood samples were collected for measurement of urea, creatinine, ALT, AST, ESR, HB, WBCs, and PLT. Relative expressions of the two biomarkers were studied: lnc-PVT1 and miR-146a. The vacutainer device was used to collect (5 mL) blood from each subject. Tubes, which had separator gels lodged between the serum layer (top) and the packed cells, contained the collected blood samples and were left to clot for a total of 15 min before being centrifuged at 4000× g for 10 min. After isolation from the clotted whole blood, the serum was stored at −80 • C until required.

Extraction of RNA
The expression of lncRNAs was evaluated via real-time PCR. With the QIAzol lysis reagent, RNA was extracted using the miRNeasy extraction kit (Qiagen, Valencia, CA, USA). The RNA concentration was determined via the NanoDrop2000 (Thermo Scientific, Wilmington, NC, USA). The extracted RNA was then stored at a temperature of −80 • C.
RT-PCR was conducted using 20 µL reaction mixtures, which were composed of a 10 µL master mix, 2.5 µL of diluted cDNA, 1 µL readymade assay primer, and 5.5 µL RNAase-free water by Rotor-Gene Q System. The conditions of the PCR were as follows: initially at 95 • C for 10 min, followed by a total of 45 cycles at 95 • C for 15 s, and finally 60 • C for 60 s. Relative to the internal control (2 −∆Ct ), the gene expression was calculated.
To determine the specificity of the RT-PCR reactions, a melt curve analysis was performed. Using 2 −∆∆Ct for relative quantification, the fold change was determined.

Study's Outcomes
The main parameter was fold change in the two studied lncRNAs, lnc-PVT1 and miR146a, and their diagnostic utilities in RA and OA patients. The secondary outcomes included the association between the fold changes in the studied biomarkers and the clinical activity of the diseases, clinical presentation, and treatment modality.

Statistical Analysis
All data was gathered and analyzed using the Statistical Package of Social Science (SPSS) software version (22) (SPSS Inc., Chicago, IL, USA). For qualitative data, the Chisquare test (for two or more groups) was used, and the data were presented in the form of numbers and percentages. An independent t-test (for two independent groups) and oneway ANOVA (for two or more independent groups) were used for quantitative information, which was presented in the form of standard deviation. Quantitative data included in this study was first tested for normality by the One-Sample Kolmogorov-Smirnov test in each study group, and then inferential statistical tests were selected. A bivariate Pearson correlation test was used to determine the association between variables. The specificity and sensitivity of the studied variable were evaluated using a ROC curve analysis. Statistical significance was considered if p-values were less than 0.05.

Results
The studied subjects were categorized into three groups: RA patients, OA patients, and HCs subjects. Each group is composed of 40 individuals. The mean age was 38.4 ± 10.1, 53.5 ± 6.7, and 52.8 ± 8.4 for the RA, OA, and HCs groups, respectively. The mean BMI for each group was 30.4 ± 5.3 kg/m 2 , 31.04 ± 4.7 kg/m 2 , 27.8 ± 5.1 kg/m 2 , respectively. The percentage of females in each group was 87.5% (RA), 75% (OA), and 82.5% (controls) ( Table 1). The laboratory investigations of different groups are illustrated in Table 2.

Expression Levels of lnc-PVT1 and miR-146a among Cases
Lnc-PVT1 was elevated in patients with RA compared to the OA group, with a fold change median 2.62 and 0.22, respectively (p = 0.001), and compared to HCs with an increase of 2 fold (Table 3). Overexpressed miR-146a marker fold change was demonstrated between both the RA and OA groups and controls but no variation between the RA and OA groups ( Table 3, Figures 1 and 2). Table 3. Comparisons of markers in different study groups. The table illustrates that there was a statistically significant difference with p-value < 0.05 between study groups as regards the lnc-PVT1 marker, with a higher fold among the RA group and a lower fold among the OA group. In addition, there was a statistically significant higher miR146a marker fold change with a p-value < 0.05 between both the RA and OA groups and HCs but no difference between the OA and RA groups.    miR-146a 4.6 1.9 6.9 10.6 1 0.2 a : significant difference between OS and control, b : significant difference between RA significant difference between OS and RA.

Correlation between lnc-PVT1 and miR-146a Markers with Disease Activity among RA Group
All patients were subjected to assessment of (DAS-28), (SJC), (ESR), and rheumatoid factor. As regards imaging, antero-posterior plain radiographs of both hands and wrists were done to evaluate the level of joint space narrowing (JSN), erosion, and osteopenia (Table 4). A positive correlation was found between the lnc-PVT1 marker and SJC among the RA group. However, there was no association between miR-146a and other variables ( Table 5).

Correlation between PVT1 and miR146a Markers with Demographic and Laboratory Investigations among RA Group
There was no statistically significant correlation (p-value > 0.05) between PVT1 and miR-146a markers level with age, BMI, and all laboratory investigations among the RA group (Table 7).

Correlation between PVT1 and miR146a Markers with Demographic and Laboratory Investigations among RA Group
There was no statistically significant correlation (p-value > 0.05) between PVT1 and miR-146a markers level with age, BMI, and all laboratory investigations among the RA group (Table 7). Table 7. Correlation between PVT1 and miR-146a markers with demographic and laboratory investigations among RA group. The table describes the relationship between both biomarkers and the biochemical characteristics of RA patients. They showed no correlation between both and any of the biochemical and demographic characteristics.

Correlation between lnc-PVT1 and miR-146a Markers with Disease Severity Scales among OA Group
The mean total KOOS score was 47.3 ± 13.4, for the chair stand test. It was 11.4 ± 6.9, and 19.7 ± 8 for the mean stair climb test. Furthermore, 37.5% of the OA group had stage 2 Kellgrn-Lawrence Grading radiological Scale, 42.5% had stage 3, and 20% had stage 4 ( Table 8). No significant correlation was found between lnc-PVT1 and miR-146a markers with different disease severity scales.

Correlation between lnc-PVT1 and miR-146a Markers with Disease Severity Scales among OA Group
The mean total KOOS score was 47.3 ± 13.4, for the chair stand test. It was 11.4 ± 6.9, and 19.7 ± 8 for the mean stair climb test. Furthermore, 37.5% of the OA group had stage 2 Kellgrn-Lawrence Grading radiological Scale, 42.5% had stage 3, and 20% had stage 4 ( Table 8). No significant correlation was found between lnc-PVT1 and miR-146a markers with different disease severity scales.

Variable
Sensitivity

Correlation between PVT1 and miR146a Markers with Demographic and Laboratory Investigations among OA Group
There was no statistically significant correlation (p-value > 0.05) between PVT1 and miR-146a markers level with age, BMI, and all laboratory investigations among OA group (Table 10). Table 10. Correlation between PVT1 and miR-146a markers with demographic and laboratory investigations among OA group. The table shows the correlation between both biomarkers and the biochemical characteristics of OA patients. There was no correlation found between either biomarker and any of the biochemical or demographic characteristics.

Discussion
Several studies examined the expression of lncRNAs in RA patients and their regulatory role in inflammation and oxidative stress of synovial fibroblasts [36][37][38][39]. Many reports have revealed interactions between lncRNA and miRNA, as well as their conurbations, in the pathogenesis of rheumatoid arthritis (RA) [27,36,[40][41][42][43]. In addition, the biological significance of lncRNA in OA, as well as the mechanisms that underpin it, has also been extensively studied [44][45][46][47]. For the first time, we studied the relationship between the expression of lnc-PVT1 and its target miRNA (miR-146a) and the pathogenesis of RA compared to OA. Lnc-PVT1 located at 8q24.21, is a novel biomarker for diagnosing cancer. It is very easy to measure its level in the serum as well as the saliva and is characterized by high specificity [48]. Moreover, lnc-PVT1 exerts regulatory effects via miRNAs modulation (cytoplasmic level), on gene transcription and protein synthesis [49]. Its role in inflammatory responses has recently been discovered [20,50]. As demonstrated in septic rats, lnc-PVT1 is highly expressed in myocardial tissues. Lnc-PVT1 reduces cardiac function and modulates the levels of inflammatory cytokines via activating mitogen-activated protein kinase (MAPK)/nuclear factor (NF)-κB pathway [51], which may suggest that it is an important contributing factor in various inflammatory diseases.
To better understand the role of lncPVT1 in the pathogenesis of RA and its susceptibility, we examined lncPVT1 expression in RA patients, OA patients, and healthy controls (HCs) and used ROC curves to assess its ability to distinguish between them. We found that lnc-PVT1 expression levels were significantly higher in RA patients, than in OA patients or HCs. Moreover, lnc-PVT1 expression levels are not correlated to disease activity, so this lncRNA may be related to the pathogenesis of RA. More importantly, ROC analysis reported that the level of lnc-PVT1 in the serum of RA patients can distinguish RA patients from healthy controls with an AUC of 0.654 at a cut off value of ≥1.15 folds with a sensitivity and specificity of 62.5% and 90%, respectively. Likewise, Zhang et al., detected elevated lnc-PVT1 expression and decreased sirt6 expression within the synovial tissues of RA-FLSs rat models. Lnc-PVT1, commonly located within the nucleus, can bind to the sirt6 promoter in order to stimulate sirt6 methylation, therefore inhibiting the transcription of sirt6. Lnc-PVT1 knockdown restored the expression of sirt6 via reducing the Life 2021, 11, 1382 11 of 14 methylation of sirt6, thereby alleviating RA [20]. Its knockdown reduced the proliferation of cells and stimulated cellular apoptosis in RA-FLSs via targeting miR-145-5p [27]. Lnc-PVT1 may stimulate inflammatory responses by activation of the MAPK/NF-κB pathway or by sponging specifically targeted miRNAs, thereby facilitating the synthesis and release of inflammatory cytokines [52]. According to Tang et al., TNF-induced RA-FLS over-proliferation was suppressed by knocking down PVT1, while TNF-induced RA-FLS apoptosis was reversed. TNF-induced production of interleukin (IL)1 and IL6, as well as NF-κB activation through miR-145-5p, were also suppressed after PVT1 knockdown [27].
MiR-146a is needed for modulating the function and differentiation of both innate and adaptive immune cells. It (or its SNP) has been confirmed as a target miRNA of lnc-PVT1 in multiple diseases (prostate cancer, colon cancer) [53]. Based on this information, we hypothesized that there may be a correlation between miR-146a and lnc-PVT1, and therefore this would be associated with the progression of RA. Interestingly, our results determined that miR-146a was up-regulated in both RA and OA patients when compared to the controls, but it neither correlated with both diseases severity signs nor with lnc-PVT1 expression levels. However, Churov et al. documented that the specificity and sensitivity of a single miRNA as a biomarker of RA is generally weak, and it is better to depend on a set of several miRNAs or a mixture of miRNAs and other parameters to demonstrate an effective diagnostic tool [54]. Zhang and coworkers found that miR-146a exacerbated pro-inflammatory cytokines by decreasing cartilage matrix-associated gene expression. MiR-146a modulates cartilage homeostasis by targeting calcium/calmodulin-dependent protein kinase II delta (Camk2d) and protein phosphatase 3, regulatory subunit B, beta isoform (Ppp3r2, also known as calcineurin B, type II). Their findings indicate that miR-146a plays a vital role in cartilage homeostasis [53]. Interestingly, MiR-146a expression was shown to be higher in CD4+ T cells, and it was found to be positively linked with TNF and IL-17 levels. The increased expression of miR-146a may contribute to RA inflammatory pathways by reducing T-cell death and increasing IL-17 cell development, respectively. Additionally, the expression of miR-146a in CD4+ T cells is inversely correlated with Fas-associated factor 1 (FAS-1), which controls T-cell death [54].
The current work found that miR-146a can be accurate in the diagnosis of patients with RA with an AUC of 0.836 at a cut-off value of ≥1.433 folds with a sensitivity and specificity of 82.5% and 100%, respectively. Therefore, it can distinguish patients from healthy controls. Different studies documented elevated expression levels of miR-146a in different sample types, such as synovial fluid, tissues, and fibroblasts, signifying its role as a potential biomarker for RA diagnosis [55][56][57]. This study used the least invasive procedure to detect the expression levels, as it can be applied to a patient's diagnosis as well as follow-up. Regarding the correlation between the level of miR-146a expression and the disease course, we demonstrated that there was no correlation between the expression levels and DAS28, ESR, and RF. Similarly, Li et al. did not find an association between miR-146a expression levels and DAI in both peripheral blood and synovial fluid [58].
Regarding the OA patients, no association was detected between miR-146a and different disease severity scales such as total KOOS score, chair stand test, stair climb test, and Kellgrn-Lawrence grading radiological scale. The small number of participants included in the study was considered a limitation. However, our results, which highlighted the value of lnc-PVT1 and miR-146a as biomarkers for RA and OA, could be a beginning for more studies on a wider scale to understand the exact role of these biomarkers in joint disorders and the ability to use them as a therapeutic target.

Conclusions
In conclusion, our study demonstrated that lnc-PVT1 and miR-146a could be considered as promising biomarkers for the diagnosis of RA and OA and may have an essential role as therapeutic targets in the future. However, no association was detected between both biomarkers and the pathogenesis of RA and OA. Further studies are required to confirm the current observations and to highlight the possibility of using them as therapeutic Life 2021, 11, 1382 12 of 14 targets in RA and OA. However, the current study has a limitation due to the small number of participating subjects.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/life11121382/s1, Figure S1: A schematic description of the study outline for the participant groups. (a) refers to a recapitulative figure describing the clinical examination, laboratory investigations, and biomarkers investigation carried out on the RA group, (b). refers to a recapitulative figure describing the clinical examination, laboratory investigations, and biomarkers investigation carried out on the OA group.