m⁶A-mediated lncRNA MAPKAPK5-AS1 induces apoptosis and suppresses inflammation via regulating miR-146a-3p/SIRT1/NF-κB axis in rheumatoid arthritis

ABSTRACT

To investigate the role of m⁶A-mediated lncRNA MAPKAPK5-AS1 (MK5-AS1) in rheumatoid arthritis fibroblast-like synoviocytes (RA-FLSs) and its underlying molecular mechanism. RT-qPCR, western blot, flow cytometry (FCM), and enzyme-linked immunosorbent assay (ELISA) were utilized for evaluating inflammation and apoptosis. Next, RIP, RNA pull-down, dual-luciferase reporter gene assay, and a series of rescue experiments were performed to explore the regulatory mechanisms of MK5-AS1 and its sponge-like action in RA-FLSs. The regulatory relationships between MK5-AS1 and WTAP were explored using the MeRIP-qPCR assay and RT-qPCR. Finally, the critical RNAs in the ceRNA axis were verified in the clinical cohort. MK5-AS1 was poorly expressed and miR-146a-3p was overexpressed in co-cultured RA-FLSs. MK5-AS1 overexpression could inhibit inflammatory responses and promote cell apoptosis in the co-cultured RA-FLSs. MK5-AS1 bound to miR-146a-3p to target SIRT1, thereby affecting inflammatory responses and cell apoptosis in the co-cultured RA-FLSs. SIRT1 knockdown or miR-146a-3p overexpression reversed the impacts of MK5-AS1 overexpression on co-cultured RA-FLSs inflammation and apoptosis. Moreover, WTAP was downregulated, and induced the inhibition of MK5-AS1 by promoting its RNA transcript stability. Clinically, MK5-AS1 was downregulated in RA-PBMCS and correlated with the clinical characteristics of RA. Our study elucidated that m⁶A-mediated MK5-AS1 sequestered miR-146a-3p to suppress SIRT1 expression in co-cultured RA-FLSs, thus providing a new insight into the treatment of rheumatoid arthritis.

1. Introduction

Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by systemic inflammation, persistent synovitis, and joint destruction [1]. RA mainly occurs in people aged 20–50 years and affects approximately 1% of the world population; the incidence rate among females is 2–3 times higher than that among males [2,3]. It is believed that RA development and progression may be attributed to both genetic and environmental factors [4]. Unfortunately, without timely diagnosis and treatment, RA may eventually result in joint damage, disability, reduced quality of life, cardiovascular events, and other comorbidities [5]. Therefore, it’s essential to uncover the underlying molecular mechanism of RA, aiming to explore effective treatment strategies for RA.

As evidenced previously, the excessive release of pro-inflammatory factors from peripheral blood mononuclear cells (PBMCs) (including macrophages, natural killer cells, dendritic cells, lymphocytes, mast cells, and T cells) disrupts the balance of the immune system, thereby accelerating RA development [6]. Fibroblast-like synoviocytes (FLSs), accepted as key effector cells in RA pathogenesis, play crucial roles in inflammation development and maintenance [7]. It has been reported that insufficient apoptosis and abnormal proliferation of FLSs are observed in RA, and excessive FLSs are tightly implicated in the various pathological processes of RA [8]. The activation of FLSs could trigger the release of pro-inflammatory factors, thereby disrupting the inflammatory balance [9]. Furthermore, it has been shown that inflammatory cytokines secreted by the immune cells could interact with FLSs and subsequently accelerate RA progression [10].

Burgeoning studies have shown that non-coding RNAs, particularly long non-coding RNAs (lncRNAs), are involved in the pathogenesis of multiple diseases (including RA) [11]. LncRNAs are defined as a type of RNA molecule with more than 200 nucleotides in length [12]. LncRNAs have been proven to be involved in various biological processes through multiple mechanisms, such as RNA interaction, protein interaction, transcription, or splicing regulation [13]. Importantly, recent studies have demonstrated that several dysregulated lncRNAs contribute to the inflammatory response and cell apoptosis in RA [14,15]. Moreover, lncRNAs have been largely suggested to exert effects on diverse diseases through the function of competing for endogenous RNA (ceRNA) via sponging microRNAs (miRs) to regulate mRNA expression [16]. For example, lncRNA SNHG1 promotes TERT expression via sponging miR-18b-5p in breast cancer [17]. LncRNA MAPKAPK5-AS1 (MK5-AS1), a novel lncRNA, has been found dysregulated in several human cancers. For instance, MK5-AS1 promotes colorectal cancer progression via cis-regulating the nearby gene MK5 and acting as a let-7f-1-3p sponge [18]. Additionally, MK5-AS1 promotes the proliferation and migration of thyroid cancer cell lines via targeting miR-519e-5p/YWHAH [19]. As has been pointed out previously, apoptosis-related MK5-AS1 is dysregulated in PBMCs of RA patients (RA-PBMCs) through RNA sequencing [20]. Nevertheless, the specific function and mechanism of MK5-AS1 in RA have not been fully reported, especially regarding its role as a ceRNA in regulating apoptosis and inflammation.

N6-methyladenosine (m⁶A) RNA methylation is the most abundant epigenetic modification in eukaryotic mRNA, which also shows presence in lncRNA, circRNA, miR, tRNA, rRNA, and snoRNA [21,22]. As a dynamic and reversible process, m⁶A RNA modification is coordinated by methyltransferase complex (m⁶A “writer”, such as METTL3, METTL14, and WTAP), demethylases (m⁶A “eraser”, such as FTO and ALKBH5), and m⁶A-binding proteins (m⁶A “reader”, such as YTH domain proteins and IGF2BP) [23,24]. Accumulating studies have demonstrated that m⁶A modification plays a vital role in RNA splicing, stability, transport, and translation [25,26]. The dysregulations of METTL3, METTL14, and WTAP have been found in liver cancer [27], lung cancer [28], and osteosarcoma [29]. Moreover, FTO and ALKBH5 act as oncogenes in various human malignancies [30]. YTHDF1–3 exerts effects on RNA translation initiation and protein synthesis via binding to initiation factors [31]. Nevertheless, little is known about the role of m⁶A modification in lncRNAs.

The modulation disorder of RNA methylation contributes to several pathological bone diseases including osteoporosis, osteoarthritis and RA. So far, the studies about m6A methylation in RA were rarely reported [32]. Sha Wu et al. demonstrated that m6A-regulated methyltransferase, demethylase, and binding protein genes are altered in the synovial tissues of patients with RA [33]. For example, METTL3-related m6A modification attenuates inflammation through the NF-κB pathway in RA, helping to recognize the pathogenesis of RA [34].

In the present study, we investigated the effects of m⁶A-mediated lncRNA MK5-AS1 methylation on rheumatoid arthritis fibroblast-like synoviocytes (RA-FLSs), and the underlying molecular axis miR-146a-3p/SIRT1/NF-κB was also explored. We hope this study could provide novel insights into developing potential molecular targets for RA treatment.

2. Materials and methods

2.1. Ethics statement

Our clinical sample collection was performed following the guidelines described in the Declaration of Helsinki for biomedical research involving human subjects. This study was approved by the Research Ethics Committee of the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine (Ethics approval number: 2019AH–12). Written informed consent was obtained from all participants before sample and data collection.

2.2. Clinical cohorts

A total of 30 RA patients (4 men and 26 women, aged 32–63 years, median age of 50.63 years) were enrolled in the Department of Rheumatology, the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine from July 2020 to October 2020. Meanwhile, 30 healthy subjects (3 men and 27 women, aged 30–78 years, median age of 52.83 years) in the physical examination center were included as the healthy controls (HCs). The inclusion criteria were determined concerning our previous studies [35]. All the subjects were included in a validation cohort to detect the expression levels of WTAP/MK5-AS1/miR-146a-3p/SIRT1 using reverse transcription-quantitative polymerase chain reaction (RT-qPCR), followed by the assessment of the application of ceRNA as a potential biomarker for RA diagnosis. The primary clinical observation indicators and self-perception of patient scores were evaluated concerning our previous study [20].

2.3. PBMC isolation

The blood samples were taken from HCs and RA patients, and fresh human PBMCs were isolated through Ficoll gradient centrifugation. Next, the isolated PMBCs were cultured in RPMI 1640 containing 10% fetal bovine serum (FBS). Overall blood from 20 different donors was used.

2.4. Establishment of a co-cultivation model of RA-PBMCs and FLSs in RA patients (RA-FLSs)

Primary human FLSs (Beijing Beina Chuanglian Biotechnology Institute, Beijing, China) were routinely monitored for mycoplasma contamination and STR profiled for cell line authentication. RA-FLSs were cultured (37°C, 5% CO₂) in 10% FBS-contained Dulbecco’s modified Eagle’s medium (DMEM) in a humid environment. Tumor necrosis factor-α (TNF-α) (10 ng/mL) (Sigma-Aldrich, St. Louis, USA) was used to stimulate RA-FLSs. RA-FLSs from 3–5 passages were used for the experiments.

RA-FLSs were co-cultured with RA-PBMCs at the ratio of 1: 1, 1: 2.5, 1: 5, and 1: 10 for 12 h, 24 h, and 48 h respectively. Cell viability was examined using the cell count kit-8 (CCK-8) assay. The expression of inflammation cytokines was detected using an enzyme-linked immunosorbent assay (ELISA).

2.5. Cell transfection

For cell transfection, MK5-AS1 overexpression plasmids (pcDNA3.1-MK5-AS1) and its small interfering RNA (si-MK5-AS1), miR-146a-3p mimics, inhibitor, sirtuin-1 (SIRT1) small interfering (si-SIRT1), and the corresponding negative controls (NC) were supplied by GenePharma (Shanghai, China). All the sequences are displayed in Supplement Table 1. Firstly, RA-FLSs were cultured in 6-well or 96-well plates 12 h before transfection following different experiment needs. When growing to 70%-80% confluence, RA-FLSs were transiently transfected using Lipofectamine 2000 (Invitrogen Inc., Carlsbad, CA, USA), with the transfection efficiency detected using RT-qPCR, followed by 48-h incubation (37°C, 5% CO₂). After that, the samples were collected at different time points for subsequent experiments.

Table 1.

Clinical characteristics of RA patients and HCs.

Variables	RA (n=30)	HC (n=30)	Normal range	P value
Gender(M/F)	4/26	3/27	NA	0.162
Age(year)	50.63±6.85	52.83±12.47	NA	0.370
Duration of disease (year)	13.50 (8.00, 19.25)	NA	NA	NA
SJC	14.87±3.27	NA	NA	NA
TJC	16.00 (12.00, 18.00)	NA	NA	NA
ESR (mm/h)	38.50 (26.75, 51.00)	3.00 (2.75, 6.25)	(2, 12)	<0.001
hs-CRP (mg/L)	5.57 (1.93, 9.22)	0.52±0.04	<1	<0.001
RF (U/ml)	84.80 (36.58, 171.53)	8.80±1.06	(0, 14)	<0.001
CCP (U/ml)	107.45 (46.75, 197.25)	2.19±0.75	<4	<0.001
IGA (g/L)	3.26±1.54	2.06±0.84	(0.7, 4)	0.003
IGG (g/L)	12.98±3.39	7.84±3.96	(7, 16)	<0.001
IGM (g/L)	1.36±0.61	1.27±0.26	(0.4, 2.5)	0.641
C3 (g/L)	1.17 (1.03, 1.31)	1.07±0.17	(0.9, 1.8)	0.047
C4 (g/L)	1.36±0.61	0.19 (0.09, 0.29)	(0.1, 0.4)	<0.001
DAS28 score	6.34±0.50	NA	NA	NA
VAS score	6.80±1.04	0.89±0.11	NA	<0.001
SAS score	58.19±4.93	30.57±2.81	NA	<0.001
SDS score	58.25 (53.50, 65.00)	32.35±3.41	NA	<0.001
PF score	60.00 (43.75, 75.00)	62.76±5.65	NA	0.574
RP score	25.00 (0.00, 56.25)	30.48±4.58	NA	0.270
BP score	52.00 (28.74, 61.99)	74.81±5.06	NA	<0.001
GH score	48.80±19.55	58.76±7.08	NA	0.023
VT score	56.50±16.87	43.23±4.99	NA	0.001
SF score	62.50 (50.00, 75.00)	60.63±4.63	NA	<0.001
RE score	33.33 (0.00, 66.66)	66.67 (33.33, 66.67)	NA	0.003
MH score	56.00 (52.00, 68.00)	65.57±3.37	NA	0.007
Symptoms and signs score	9.00 (6.00, 18.00)	NA	NA	NA
Spleen deficiency and dampness syndrome score	9.50 (5.75, 17.00)	NA	NA	NA

HC: healthy control; SJC: swollen joint count; TJC: tender joint count; ESR: erythrocyte sedimentation rate; hs-CRP: high-sensitivity C-reactive protein; RF: rheumatoid factor; CCP: citrullinated peptide antibodies; IGA: immunoglobulin A; IGG: Immunoglobulin G; IGM: Immunoglobulin M; C3: complement C3; C4: complement C4; DAS28: disease activity score of 28 joints; VAS: visual analog scale; SAS: self-rating anxiety scale; SDS: self-rating depression scale; PF: physical functioning; RP: role-physical; BP: body pain; GH: general health; VT: vitality; SF: social functioning; RE: role-remotional; MH: mental health.

2.6. RNA fluorescence in situ hybridization (FISH)

The sequence of the oligonucleotide-modified probe MK5-AS1 was synthesized by GenePharma. The cells were placed on slides and fixed in 4% paraformaldehyde for 5 min, followed by phosphate-buffered saline (PBS) washing. Next, cells were washed and treated with RNase A (15 min, 37°C). After dehydration through ethanol series, probes were hybridized (overnight, 37°C). Following 2 washes with 50% formamide/0.5×, the cells were incubated (30 min, 37°C) with Alexa Fluor 488 reagent (Proteintech, Wuhan, China). Images were obtained using a fluorescence microscope (Olympus Optical Co., Ltd, Tokyo, Japan). The probe sequence is as follows: MK5-AS1: 5’-CACGACGCCAAACGGCTACCTTGTAAAGACGAAA (ttt CATCATCAT ACATCATCAT)30–3’.

2.7. Dual-luciferase reporter gene assay

The wild-type (MK5-AS1-WT and SIRT1-WT containing binding sites for miR-146a-3p) and mutant plasmids (MK5-AS1-MUT and SIRT1-MUT) were consolidated into the pGL3 promoter vector (GenePharma). HEK293T cells were cultured on 96-well plates and then co-transfected with wild-type/mutant luciferase plasmids and miR-146a-3p/control miRNA. After that, the luciferase activity was determined using the Dual-Luciferase Reporter Assay Kit (Promega Corp., Madison, Wisconsin, USA). All the experiments were carried out at least three times.

2.8. RNA pull-down assay

RNA pull-down assay was performed using the Pierce™ Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific Inc., Waltham, MA, USA). Briefly, 3ʹ-biotinylated miR-146a-3p (Bio-miR-146a-3p) or Bio-miR-NC was transfected into HEK293T cells. After 48 h, streptavidin-coated magnetic beads (Invitrogen) were added to the cell lysates to pull down the biotin-coupled RNA complex. Finally, the enrichment of MK5-AS1 and miR-146a-3pin bound fractions was detected using RT-qPCR.

2.9. RNA immunoprecipitation (RIP) assay

The relationship between MK5-AS1 and miR-146a-3p was determined using the Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore, Billerica, MA, USA). RIP assay was performed using the control IgG and human Ago2 antibodies in RIP buffer. Next, the coprecipitated RNAs were collected for cDNA synthesis and evaluated using RT-qPCR.

2.10. RNA extraction and RT-qPCR

Total RNA was isolated from RA-FLSs using TRIzol reagent (Invitrogen) following the manufacturer’s protocols. Next, RNA was reverse-transcribed to cDNA using a Reverse Transcription Kit (Takara Bio Inc., Kyoto, Japan). RT-qPCR analyses were performed using TB Green™ Premix Ex Taq™II (TaKaRa). The reaction conditions were pre-denaturation at 95°C for 1 min, followed by 40 cycles of 95°C for 20 s and 60°C for 1 min. All the primers were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). The primer sequences are listed in Supplementary Table S1. Data were normalized using the 2^−ΔΔCT method relative to U6 and β-actin.

2.11. ELISA

The supernatants of the treated RA-FLSs were collected and centrifuged (1500rpm, 10 min, 4°C). The concentrations of human interleukin-4 (IL-4), IL-10, IL-6, and IL-8 in the supernatants were determined using the corresponding ELISA kits as per the manufacturer’s instructions (Quanzhou Ruixin Bio-Tech Co., Ltd).

2.12. Flow cytometry (FCM)

Cell apoptosis was determined using Annexin V-FITC/PI Apoptosis Detection Kit (Sangon Biotech). Briefly, the cells were harvested through centrifugation (1500rpm, 6 min, 25°C). Next, cells were slightly resuspended with 1 × binding buffer and then added with FITC-Annexin V and PI at 37°C for 15 min away from light. Finally, the number of apoptotic cells was evaluated using a flow cytometer (BioRad, CA, USA).

2.13. Western Blot (WB)

Cells were washed twice with PBS and harvested in 100 µL lysis buffer through scraping. Next, the lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for separation, followed by the transferring to polyvinylidene difluoride (PVDF) membranes. After two washes with 1× TBST, the PVDF membranes were blocked in 5% nonfat milk (2 h, room temperature), and incubated (overnight, 4°C) with primary antibodies anti-Bax (1:1000; ab32503, Abcam Inc., Cambridge, MA, USA), anti-Bcl-2 (1:1000; ab32124, Abcam), anti-caspase3 (1:5000; ab32351, Abcam), anti-caspase8 (1:1000; AF6442, Affinity), METTL3 (1: 1000; 382974, ZENBIO), METTL14 (1: 1000; A8530, ABclonal), and WTAP (1: 1000; DF3282, Affinity). After TBST washing, the membranes were incubated (2 h, room temperature) with goat anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (1:3000; Abcam). The protein level was measured using an enhanced chemiluminescence detection kit (Thermo Fisher). Blots were quantified using ImageJ analysis software. All the experiments were performed three times.

2.14. m⁶A RNA methylation quantification

The total m⁶A level of the extracted RNA was detected using an m⁶A RNA Methylation Quantification Kit (Abcam) following the manufacturer’s instructions. Briefly, total RNA was bound to wells using the RNA high-binding solution. m⁶A was detected using capture and detection antibodies. After the addition of the color-developing reagent, the m⁶A level was then quantified by measuring the absorbance at 450 nm.

2.15. m⁶A immunoprecipitation qPCR (MeRIP-qPCR)

m⁶A modification on MK5-AS1 was determined using the Magna MeRIP m⁶A Kit (Millipore) according to the manufacturer’s instructions. Briefly, Poly-A mRNA was first isolated using the Dynabeads® mRNA Purification Kit (Invitrogen). Pierce protein A/G beads were washed three times with a specific buffer, followed by incubation (2 h, 4°C) with m⁶A primary antibody with rotation. Next, beads were washed with buffer and 100 ng of poly(A) + RNA, and 1 × immunoprecipitation buffer supplemented with RNase inhibitors. Afterward, the methylated mRNAs were precipitated by adding 2.5 volumes of 100% ethanol with glycogen and overnight incubation (−80°C). The m⁶A enrichment was finally determined using qPCR analysis.

2.16. Statistical analysis

Statistical analysis was performed using GraphPad Prism version 8.1 (GraphPad Software). The quantitative data were expressed as medians with interquartile range (IQR) or means  ±  SEM. The differences between the groups were analyzed using Student’s t-test or nonparametric for continuous variables and the Chi-square test for categorical variables. Comparisons between multiple-group means were performed using one-way analysis of variance (one-way ANOVA). A value of p < 0.05 represented a statistically significant difference. The correlation between two variables was analyzed using Spearman’s correlation analysis.

3. Results

3.1. RA-PBMCs improved the proliferation and inflammatory response of RA-FLSs

To ascertain the actions of PBMCs on the proliferation and inflammatory response of RA-FLSs, we extracted PBMCs from RA patients, which were subsequently co-cultured with RA-FLSs. According to CCK-8 assay results, a ratio of 2.5: 1 (RA-PBMCs: RA-FLSs) was the most optimal for RA-PBMCs to stimulate RA-FLSs, and RA-FLSs showed peak viability at the 48 h (Figure 1(a)). Therefore, a ratio of 2.5: 1 was chosen for the subsequent experiments. It was found that both RA-PBMCs and TNF-α can stimulate RA-FLS proliferation, with RA-PBMCs showing much less robust stimulation than TNF-α. However, RA-PBMCs combined with TNF-α exhibited slightly better effects on RA-FLS proliferation than TNF-α alone (Figure 1(b)). Next, ELISA was performed to assess the effect of RA-PBMCs on the inflammatory response of RA-FLSs. The results indicated that RA-PBMCs combined with TNF-α stimulation could elevate pro-inflammatory cytokine IL-8 expression (Figure 1(c)), and downregulate anti-inflammatory cytokine IL-10 expression (Figure 1(d)).

Figure 1.

RA-PBMCs slightly improve the proliferative and inflammatory response potentials of RA-FLSs. (a) RA-PBMCs and RA-FLSs were co-cultured at a ratio of 0:1, 1:1, 2.5:1, 5:1, and 10:1(n=4). (b) The CCK-8 assay showed that RA-PBMCs affected RA-FLS proliferation (n=4). (c/d) the ELISA assay showed that RA-PBMCs affected RA-FLS inflammatory response (n=6). Data were represented as the mean ± SD. vs. RA-FLSs, **p < 0.01, ***p < 0.001. vs. TNF-α+RA-FLSs, ▴p < 0.05, ▴▴▴p < 0.001. Statistical analysis was performed using the one-way ANOVA.

3.2. MK5-AS1 was downregulated in the co-cultured RA-FLSs

MAPKAPK5 maps to 111,839,764–111,842,902 in Chromosome 12 (Figure 2(a)). As shown in FISH assay results, MK5-AS1 was mainly distributed in the cytoplasm (Figure 2(b)). The TNF-α-stimulated group served as the control group, and the group stimulated by RA-PBMCs and TNF-α served as the co-cultured group (Co-C) (Figure 2(c)). To evaluate MK5-AS1 function in RA-FLSs, we constructed MK5-AS1 overexpression plasmid (OV-MK5-AS1), and three siRNAs (#1, #2, and #3) targeting the junction sites of MK5-AS1. First, it was found that MK5-AS1 was notably downregulated in the Co-C group compared with that in the C group (Figure 2(c)). MK5-AS1 expression was remarkably increased or decreased after transfection of MK5-AS1 overexpression plasmid or siRNAs targeting MK5-AS1, respectively; si-MK5-AS1 #2 exhibited the best transfection efficiency (Figure 2(d,e)).

Figure 2.

MK5-AS1 was downregulated in RA-FLSs. (a) MK5-AS1 is located on chromosome 12q24.12. (b) The FISH assay detected the subcellular localization of MK5-AS1 (magnification 400 ×). (c) The expression of MK5-AS1 in the co-cultured cells was detected using RT-qPCR. (d) The overexpression efficiency of MK5-AS1 was detected using RT-qPCR. (e) The depletion efficiency of MK5-AS1 was detected using RT-qPCR. Data were represented as mean ± SD. ns, not significant. vs. control group, ***p < 0.001. The experiments were repeated independently at least three times.

3.3. LncRNA MK5-AS1 overexpression inhibited inflammation and promoted apoptosis in the co-cultured RA-FLSs

To further investigate the role of MK5-AS1 in cell inflammation and apoptosis, we performed a series of experiments to evaluate the effect of MK5-AS1 knockdown and overexpression on the co-cultured cells. As revealed by ELISA results, MK5-AS1 knockdown dramatically increased pro-inflammatory cytokines (IL-6 and IL-8) expressions and decreased anti-inflammatory cytokines (IL-4 and IL-10) expressions, whereas MK5-AS1 overexpression played the opposite role (Figure 3(a-d)). Next, the role of MK5-AS1 in the apoptosis of the co-cultured cells was determined. As shown by WB results, MK5-AS1 overexpression significantly increased Bax, caspase3/cleaved caspase3, and caspase8/cleaved caspase8 protein levels but reduced Bcl-2 level, which were all reversed after MK5-AS1 knockdown in si-MK5-AS1-transfected co-cultured cells (Figure 3(e)). Simultaneously, MK5-AS1 overexpression was found to significantly promote cell apoptosis in the co-cultured cells, whereas MK5-AS1 knockdown caused the opposite results (Figure 3(f)). Taken together, these results indicated that MK5-AS1 overexpression inhibited inflammation and promoted apoptosis in the co-cultured cells.

Figure 3.

LncRNA MK5-AS1 overexpression inhibited inflammation and promoted apoptosis in the co-cultured cells. (a-d) ELISA assay was utilized to measure the expression levels of inflammatory cytokines after MK5-AS1 overexpression and knockdown (n=6); (e) the changes in apoptosis-related protein levels were measured using WB after MK5-AS1 overexpression and knockdown (n=3); (f) FCM experiment was adopted to detect cell apoptosis after MK5-AS1 overexpression and knockdown(n=3). Data were represented as mean ± SD. ns, not significant. vs. Control group, ***p < 0.001. vs. cells transfected with si-NC, ^#p < 0.05, ^##p < 0.01, ^###p < 0.001. vs. cells transfected with OV-NC, ▴p < 0.05, ▴▴p < 0.01, ▴▴▴p < 0.001. Statistical analysis was performed using the one-way ANOVA.

3.4. LncRNA MK5-AS1 served as a sponge for miR-146a-3p

Thereafter, we planned to explore the downstream mechanism of MK5-AS1 in RA-FLSs. Extensive studies have confirmed that lncRNAs can regulate gene expression via their miRNA responsive elements (MREs), which locate in the cytoplasm and are tightly implicated in the ceRNA network. In this study, RNA pull-down experiments were conducted to select the eligible miRNAs in RA. According to the results, miR-146a-3p was highly enriched in the pull-down compounds of the MK5-AS1 biotin probe (Figure 4(a)). miRs bind to their target genes and exert effects on translational repression or RNA degradation mainly in an AGO2-dependent manner. Therefore, the RIP experiment was performed using anti-AGO2 antibodies to confirm the direct binding between miR-146a-3p and MK5-AS1. It was observed that MK5-AS1 and miR-146a-3p were enriched in the AGO2 immunoprecipitates relative to those in the IgG (Figure 4(b)). Furthermore, a dual-luciferase reporter gene assay was performed to confirm the binding between MK5-AS1 and miR-146a-3p. The results indicated that miR-146a-3p mimics notably decreased the luciferase activity of the wild-type MK5-AS1 vector, with no significant change found in the luciferase activity of the mutant MK5-AS1 vector (Figure 4(c)). Additionally, miR-146a-3p expression was remarkably increased or decreased after MK5-AS1 knockdown or overexpression, separately, in the co-cultured cells (Figure 4(d)). Subsequently, miR-146a-3p expression in the co-cultured cells and its influence on MK5-AS1 were investigated. It was found that the miR-146a-3p expression was significantly increased and MK5-AS1 expression was notably decreased after the transfection of miR mimics, while the opposite results were obtained after miR-inhibitor transfection (Figure 4(e,f)). From all the above, MK5-AS1 could directly interact with the miR-146a-3p in RA-FLSs.

Figure 4.

LncRNA MK5-AS1 serves as a sponge for miR-146a-3p in co-cultured cells. (a-c) RNA-pull down, RIP, and dual-luciferase reporter gene assay were implemented to verify the binding between MK5-AS1 and miR-146a-3p (n = 6). (d) RT-qPCR was utilized to detect miR-146a-3p expression after MK5-AS1 overexpression and knockdown (n = 6). (e,f) RT-qPCR was implemented to detect the expression of MK5-AS1 and miR-146a-3p after miR-146a-3p overexpression and knockdown (n = 6). Data were represented as mean ± SD. ns, not significant. vs. cells transfected with mimic NC, ***p < 0.001. vs. cells transfected with inhibitor NC, ▴▴p < 0.05, ▴▴▴p < 0.001. Statistical analysis was performed using the two-sample independent t test and one-way ANOVA.

3.5. SIRT1 was a downstream target of miR-146a-3p

Subsequently, we shifted to investigating the downstream mechanism. As has been pointed out previously, miR-146a-3p is the target miRNA of SIRT1 [36]. As expected, in the present study, the luciferase activity of SIRT1-WT was effectively reduced by miR-146a-3p mimics (Figure 5(a)). More importantly, MK5-AS1 and SIRT1 were all captured by anti-AGO2, which indicated the coexistence of MK5-AS1 and SIRT1 in RNA-induced silencing complexes (RISCs) (Figure 5(b)). Next, miR-146a-3p effects on SIRT1 expression and the NF-κB pathway were analyzed. As revealed by WB results, miR-146a-3p overexpression increased p-p65 and p-p50 protein levels, and miR-146a-3p inhibition caused opposite results (Figure 5(c)). Moreover, miR-146a-3p overexpression was reduced, whereas miR-146a-3p knockdown enhanced SIRT1 expression, which was further verified using RT-qPCR (Figure 5(c,d)). Taken together, SIRT1 was a direct target of miR-146a-3p.

Figure 5.

SIRT1 was a downstream target of miR-146a-3p. (a,b) Dual-luciferase reporter gene assay and RIP assay were implemented to evaluate the interactions among MK5-AS1, miR-146a-3p, and SRT1 (n = 3). (c) WB was utilized to detect the protein levels of SIRT1, p-p65, and p-p50 after transfection of miR-146a-3p mimic and inhibitor (n = 4). (d) RT-qPCR was used to detect SIRT1 after transfection of miR-146a-3p mimic and inhibitor(n = 6). Data were represented as mean ± SD. ns, not significant. vs. cells transfected with mimic NC, ***p < 0.001. vs. cells transfected with inhibitor NC, ▴▴p < 0.05, ▴▴▴p < 0.001. Statistical analysis was performed using the two-sample independent t test and one-way ANOVA.

3.6. miR-146a-3p overexpression reversed the influence of MK5-AS1 overexpression on apoptosis and inflammation in RA-FLSs

Next, functional rescued experiments were performed to explore whether MK5-AS1 regulated inflammation and apoptosis via modulating miR-146-a-3p. It was discovered that MK5-AS1 overexpression decreased inflammatory cytokine expressions and increased anti-inflammatory cytokine expressions, which were restored by miR-146a-3p overexpression (Figure 6(a-d)). Meanwhile, MK5-AS1 overexpression remarkably reduced anti-apoptotic Bcl-2 protein level and elevated pro-apoptotic Bax, caspase3/cleaved caspase3, and caspase8/cleaved caspase8 protein levels, which were all reversed by miR-146a-3p overexpression (Figure 6(e)). Likewise, miR-146a-3p upregulation could offset the promoting effect of MK5-AS1 overexpression on apoptosis in the co-cultured cells (Figure 6(f)). Therefore, MK5-AS1 overexpression induced apoptosis and suppressed inflammatory response via binding to miR-146a-3p.

Figure 6.

miR-146a-3p overexpression reversed the influence of MK5-AS1 overexpression on apoptosis and inflammation in RA-FLSs. (a-d) ELISA was utilized to detect the expression of the inflammatory cytokines after MK5-AS1 and miR-146a-35p overexpression (n = 6). (e) The apoptosis-related protein levels were examined using WB (n = 3). (f) FCM analysis was performed to measure the rate of apoptosis (n = 3). Data were represented as mean ± SD. ns, not significant. vs. cells co-transfected with OV-NC and mimic NC, ***p < 0.001. vs. cells co-transfected with OV-MK5-AS1 and miR-146a-3p mimic, ^###p < 0.001. Statistical analysis was performed using the one-way ANOVA.

3.7. SIRT1 knockdown reversed the influence of MK5-AS1 overexpression on apoptosis and inflammation in the co-cultured RA-FLSs

Subsequently, rescue experiments were implemented to evaluate whether MK5-AS1 affected RA inflammation and apoptosis via the SIRT1/NF-KB signaling. First of all, the knockdown efficiency of si‐SIRT1 in the co-cultured cells was evaluated, and si-SIRT1 #2 with the highest knockdown efficiency was chosen for the subsequent experiments (Figure 7(a)). Resultantly, MK5-AS1 overexpression-induced SIRT1 downregulation could be reversed after SIRT1 depletion (Figure 7(b)). Moreover, the effect of MK5-AS1 overexpression on the expression of inflammatory cytokines was reversed after SIRT1 inhibition (Figure 7(c-f)). Furthermore, as shown by WB and FCM analysis results, MK5-AS1 overexpression-influenced apoptotic protein expressions and MK5-AS1 overexpression-stimulated cell apoptotic ability could be reversed by SIRT1 inhibition (Figure 7(g,h)). Moreover, SIRT1 knockdown could offset the suppressing effect of MK5-AS1 overexpression on the NF-κB signaling pathway in RA (Figure 7(i)). To sum up, MK5-AS1 induced apoptosis and suppressed the inflammatory response in RA-FLSs via augmenting SIRT1.

Figure 7.

SIRT1 knockdown reversed the influence of MK5-AS1 overexpression on apoptosis and inflammation in RA-FLSs. (a) the knockdown efficiency of SIRT1 in the co-cultured cells was detected using RT-qPCR (n = 3). (b) SIRT1 expression was detected using RT-qPCR after MK5-AS1 overexpression and SIRT1 inhibition (n = 6). (c/f) inflammatory cytokine expressions in different groups were determined using ELISA (n = 6). (g)The apoptosis-related protein levels were examined using WB (n = 3). (h) FCM analysis was performed to measure the rate of apoptosis (n = 3). (i) WB was used to detect the protein levels of SIRT1, p-p65, and p-p50 (n = 4). Data were represented as mean ± SD. ns, not significant. vs. cells co-transfected with OV-NC and si-NC, ***p < 0.001. vs. cells co-transfected with OV-MK5-AS1 and si-NC, ▴▴▴p < 0.001. Statistical analysis was performed using the one-way ANOVA.

3.8. WTAP was downregulated in co-cultured RA-FLSs

Our results represented 12 RRACU m⁶A sequence motifs in the exon region based on the online bioinformatics database m⁶Avar (http://m6avar.renlab.org). Next, the total m⁶A methylation level and methylation percentage in the co-cultured cells were detected. It was found that the co-cultured cells showed a notably reduced level of total m⁶A methylation and methylation percentage (Figure 8(a,b)). m⁶A on RNA is mediated by a methyltransferase complex (METTL3, METTL14, and WTAP), and investigating this complex is crucial for advancing research on m⁶A. In the present study, the results revealed that there was no significant difference in METTL3 protein level in the co-cultured cells, while METTL14 and WTAP protein levels were remarkably decreased, with WTAP protein levels showing a more dramatic reduction (Figure 8(c)). Subsequently, WTAP overexpression plasmid (OV-WTAP) was constructed and the overexpression efficiency was confirmed using RT-qPCR and WB. As shown in Figure 8(d,e), both the mRNA expression and protein level of WTAP were dramatically increased in WTAP-overexpressed co-cultured cells. Finally, it was found that WTAP overexpression reduced MK5-AS1 expression in co-cultured cells (Figure 8(f)).

Figure 8.

WTAP was downregulated in the co-cultured RA-FLSs. (a) m⁶A level in the co-cultured cells was determined (n = 3). (b) The m⁶A methylation percentage in the co-cultured cells was detected (n = 3). (c) The protein levels of METTL3, METTL4, and WTAP were measured using WB (n = 3). (d,e) overexpression of WTAP was confirmed by RT-qPCR (n = 6) and WB (n = 4) in the co-cultured cells. (f) RT-qPCR was performed to determine WTAP effects on the expression of MK5-AS1 (n = 6). Data were represented as mean ± SD. ns, not significant. vs. Control group, ***p < 0.001. vs. cells co-transfected with OV-NC, ^###p < 0.001. Statistical analysis was performed using the two-sample independent t test and one-way ANOVA.

3.9. m⁶A modification was associated with MK5-AS1 downregulation in the co-cultured RA-FLSs

Additionally, MeRIP-qPCR was conducted to clarify whether WTAP could directly mediate the m⁶A methylation of MK5-AS1. As revealed by the results, m⁶A was lower enriched within MK5-AS1 in the co-cultured cells (Figure 9(a)). It was found that WTAP overexpression increased the m⁶A level of MK5-AS1 in the co-cultured cells (Figure 9(b)).

Figure 9.

m6A modification was associated with MK5-AS1 downregulation in the co-cultured RA-FLSs. (a) m⁶A methylation level of MK5-AS1 in the co-cultured cells was determined using MeRIP-qPCR (n = 3). (b) Changes in m⁶A-modified MK5-AS1 levels upon WTAP overexpression in the co-cultured cells were detected (n = 3). Data were represented as mean ± SD. ***p < 0.001. vs. Control group and cells co-transfected with OV-NC. Statistical analysis was performed using the two-sample independent t test.

3.10. Clinical characteristics of study subjects were compared

In the present study, a total of 60 subjects (30 RA patients and 30 HCs) were enrolled in the validation set for evaluating the effects of the WTAP/MAPKAPK5-AS1/miR-146a-3p/SIRT1 axis. The characteristics of the participants are summarized in Table 1. No significant difference between RA patients and HC was found in age or gender. There were significant differences in erythrocyte sedimentation rate (ESR), high-sensitivity C-reactive protein (hs-CRP), rheumatoid factor (RF), citrullinated peptide antibodies (CCP), immunoglobulin A (IGA), immunoglobulin G (IGG), complement C3 (C3), complement C4 (C4), visual analog scale (VAS), self-rating anxiety scale (SAS), self-rating depression scale (SDS), body pain (BP), general health (GH), vitality (VT), social functioning (SF), role-remotional (RE), and mental health (MH) between RA patients and HCs.

3.11. Expressions of the WTAP/MK5-AS1/miR-146a-3p/SIRT1 axis in PBMCs of RA patients were detected

Subsequently, the expression of WTAP, MK5-AS1, miR-146a-3p, and SIRT1 in the PBMCs of RA cases and HCs was detected. The results revealed that RA patients exhibited notably lower expressions of MK5-AS1, SIRT1, METTL3, METTL14, and WTAP, and higher miR-146a-3p expression than HCs (Figure 10(a-f)). Next, the associated area under curve (AUC) was calculated and the receiver operating characteristic (ROC) curve was plotted. As shown by the ROC curves, MK5-AS1 [0.9343 (95% CI  =  0.8693–0.9992)], miR-146a-3p [0.8617 (95% CI  =  0.7659–0.9574)], SIRT1 [0.7550 (95% CI  =  0.6284–0.8716)], METTL3 [0.8522 (95% CI  =  0.7563–0.9481)], METTL14 [0.7256 (95% CI  =  0.5973–0.8538)], and WTAP [0.7206 (95% CI  =  0.5842–0.8569)] had high AUC for RA diagnosis (Figure 10(g-m)).

Figure 10.

Expressions of the WTAP/MK5-AS1/miR-146a-3p/SIRT1 axis in PBMCs of RA patients were detected. (a) the mRNA expression of MK5-AS1 was detected. (b) The mRNA expression of miR-146a-3p was detected. (c) The mRNA expression of SIRT1 was detected. (d) The mRNA expression of METTL3 was detected. (e) The mRNA expression of METTL14 was detected. (f) The mRNA expression of WTAP was detected. (g) The ROC curve of MK5-AS1 was plotted. (h) The ROC curve of miR-146a-3p was plotted. (i) The ROC curve of SIRT1 was plotted. (j) The ROC curve of METTL3 was plotted. (k) The ROC curve of METTL14 was plotted. (l) The ROC curve of WTAP was plotted. (m) The ROC results of WTAP, MK5-AS1, miR-146a-3p, METTL3, METTL14, and SIRT1.

3.12. The correlation of clinical features with WTAP/MK5-AS1/miR-146a-3p/SIRT1 expressions in RA-PBMCs was analyzed

Subsequently, we collected the clinical features of RA and assessed their correlation with the expression of the WTAP/MK5-AS1/miR-146a-3p/SIRT1 axis. As shown in Supplementary Figure S2, MK5-AS1 expression was negatively correlated with miR-146a-3p, swollen joint count (SJC), hs-CRP, IGG, symptoms and signs score, and spleen deficiency and dampness syndrome score, and positively correlated with SIRT1, METTL14, WTAP, and RP (Figure 11(a-j)). miR-146a-3p expression was negatively correlated with SIRT1, METTL14, and WTAP (Figure 11(k-m)), and positively correlated with tender joint count (TJC), RF, DAS28, and spleen deficiency and dampness syndrome score (Figure 11(n-q)). There was a negative correlation between SIRT1 expression and ESR, SAS, and physical functioning (PF), and a positive correlation between SIRT1 expression and social functioning (SF), METTL14, and WTAP (Figure 11(r-w)). SIRT1 negatively correlated with ESR, symptoms and signs score, and spleen deficiency and dampness syndrome score (Figure 11(x-z)). However, no correlation was found between these RA clinical features and METTL3 and METTL14 mRNA expressions (data not shown).

Figure 11.

The correlation of clinical features with WTAP/MK5-AS1/miR-146a-3p/SIRT1 expressions in RA-PBMCs was analyzed. (a-j) correlation between MK5-AS1 expression and miR-146a-3p, SIRT1, METTL14, WTAP, SJC, hs-CRP, IGG, RP, symptoms and signs score, and spleen deficiency and dampness syndrome score was analyzed. (k-q) the correlation between miR-146a-3p expression and SIRT1, METTL14, WTAP, TJC, RF, DAS28 score, and spleen deficiency and dampness syndrome score was analyzed. (r-v) the correlation between SIRT1 expression and ESR, SAS, PF, SF, METTL14, and WTAP was analyzed. (x-z) the correlation between WTAP expression and ESR and spleen deficiency and dampness syndrome score was analyzed.

4. Discussion

Growing studies have evidenced that the dysregulation of lncRNAs is closely linked to the progression of various rheumatic diseases. A majority of lncRNAs have been revealed to participate in RA progression through the ceRNA mechanism. For example, LOC100912373 modulates PDK1 expression via sponging miR-17-5p to promote the proliferation of RA-FLS [37]. HIX003209 promotes inflammation via sponging miR-6089 involving the TLR4/NF-κB signaling pathway in RA [38]. In the present study, MK5-AS1, a novel lncRNAs, was found downregulated in RA, and MK5-AS1 overexpression could induce RA-FLS apoptosis and suppress the inflammatory response. The whole experimental results indicated that MK5-AS1 presented a crucial role in RA pathogenesis.

Subsequently, we further found that MK5-AS1 was distributed in the cytoplasmic portion, which may function as a miRNA “sponge”. The results demonstrated that MK5-AS1 targeted miR-146a-3p. Moreover, miR-146a-3p targeted the 3’-UTR of SIRT1 mRNA. Overall, we proved that the MK5-AS1/miR-146a-3p/SIRT1 axis regulated RA-FLS inflammation and apoptosis, thereby influencing RA progression. Consistently, multiple lncRNAs have been proposed to regulate inflammation and apoptosis in RA. For instance, lncRNA ZFAS1 regulates the proliferation, apoptosis, inflammatory response, and autophagy of RA-FLS via regulating the miR-2682-5p/ADAMTS9 axis [39]. LncRNA PVT1 knockdown suppresses RA-FLS inflammation and induces apoptosis in RA through the demethylation of SIRT6 [40]. Besides, the relationship between m⁶A methylation and MK5-AS1 was also confirmed.

miRs are defined as a group of small molecular non-coding RNAs with 18–25 nucleotides in length [41]. miR-146a has been shown to play an important role in the regulation of inflammatory innate immune responses [42]. A previous study has pointed out that miR-146a-3p suppresses the differentiation of human amniotic mesenchymal stem cells into Schwann cells via regulating ERBB2 [43]. Moreover, the ceRNA network of lncRNA-miR-146a-3p has been reported to be tightly implicated in castration-resistant prostate cancer [44]. As has been evidenced previously, miR-146a is involved in RA pathogenesis, which is mediated by the polymorphisms in miR-146a [45]. In this study, miR-146a-3p was identified as the downstream miR of MK5-AS1, and miR-146a-3p was found upregulated in RA. The MK5-AS1/miR-146a-3p axis was demonstrated to be closely related to RA pathogenesis. MK5-AS1 overexpression could induce RA-FLS apoptosis and suppress the inflammatory response, which was partially rescued by miR-146a-3p overexpression.

SIRT1, as the most widely studied member of the SIRT family, could modulate cell proliferation, apoptosis, migration, and invasion [46]. It has been disclosed that SIRT1 promotes proliferation and pro-inflammatory cytokine production in RA, which indicated that SIRT1 is involved in RA progression [47]. Additionally, as has been reported previously, SIRT1 could suppress the NF-κB signaling, thereby greatly attenuating inflammation driven by the NF-κB signaling pathway [48]. Consistently, SIRT1 was found poorly expressed in RA-FLSs, and the NF-κB signaling pathway was negatively associated with SIRT1 in this study. Finally, the rescue experiments validated that MK5-AS1 participated in RA progression via regulating the miR-146a-3p/SIRT1 axis.

WTAP acting as a critical catalyzing enzyme is responsible for m⁶A modification, which can improve methylation efficiency in nuclear speckles by translocating the METTL3-METTL14 complex to mRNA targets [49]. Mounting studies have suggested that WTAP participates in the pathogenesis of various diseases. For instance, Wei W et al. have discovered that circ0008399 could interact with WTAP to promote the assembly and activity of the m⁶A methyltransferase complex and promote cisplatin resistance in bladder cancer [50]. Liu W et al. have reported that miR-139-5p loss-mediated WTAP activation contributes to hepatocellular carcinoma progression via promoting the epithelial to mesenchymal transition [51]. Recently, WTAP has been identified as a vital m⁶A methylation gene through m⁶A-seq, showing involvement in RA onset and progression [52]. We for the first time demonstrated that WTAP was downregulated in RA in this study, and WTAP downregulation was strongly associated with the ceRNA network and clinical laboratory indexes in RA. Moreover, it was found that WTAP-mediated the m⁶A modification of MK5-AS1 and inhibited its expression. From all the above, we proved that m⁶A methylation induced MK5-AS1 downregulation in RA.

There are some limitations to this study. First, the clinical samples in the validation set were relatively small, and large-scale samples are needed for follow-up verification. Next, due to it being difficult for us to obtain the tissues from RA patients in a short time, and most RA patients are willing to accept the conservative treatment of traditional Chinese medicine, we did not obtain primary FLSs from RA patients to confirm the results in vitro. Moreover, this study is relatively preliminary research, the relationship between m⁶A “eraser” and m⁶A “reader” with MK5-AS1 is not thoroughly explored in this study, which will become the focus of our following research. Finally, the relationship of MK5-AS1 with the efficacy and prognosis of patients receiving different treatments will be evaluated in the future.

In conclusion, we identified that MK5-AS1 downregulation was tightly implicated in the co-cultured RA-FLSs, which may be regulated by WTAP-mediated m⁶A modification. MK5-AS1 could induce RA-FLSs apoptosis and suppress the inflammatory response in the co-cultured RA-FLSs via regulating the miR-146a-3p/SIRT1 axis (Figure 12). Therefore, MK5-AS1 is a potential diagnostic indicator and suitable therapeutic target in RA treatment.

Figure 12.

Schematic diagram demonstrating the molecular mechanisms underlying MK5-AS1 in RA.

Funding Statement

This work was supported by grants from the Ministry of Science and Technology National Key Research and Development Program Chinese Medicine Modernization Research Key Project (2018YFC1705204); National Nature Fund Program (82274490, 81973655, 82074373); The University Synergy Innovation Program of Anhui Province (GXXT-2020-025); Open Foundation of Key Laboratory of Xin’an Medical Ministry of Anhui University of Traditional Chinese Medicine (No.2020×ayx10); Open Foundation of Anhui Province Key Laboratory of Modern Chinese Medicine Department of Internal Medicine Application Foundation Research and Development (2021AKLMCM004); Anhui Province Major and Intractable Diseases Collaborative Research Project of Traditional Chinese and Western Medicine (Anhui Traditional Chinese Medicine Development [2021] No. 70).

Author contributions

LJ, WL, JH, and WJT conceived and designed the experiment. WJT, WX, and WJ performed most of the experiments. SY and WL contributed to the patient samples. WJT and SY analyzed the statistical analyses. XL and FYY guidelined to how to proceed with clinical data mining. WJT wrote the manuscript. All authors revised the manuscript. All authors contributed to the article and approved the submitted version.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

All relevant data and materials are stored at the Anhui Province Key Laboratory of Modern Chinese Medicine Department of Internal Medicine Application Foundation Research and Development and can be obtained from the first author and corresponding author.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15384101.2024.2302281