The comprehensive clinical data of 326 discharged RA patients were collected from the electronic health recordsystem of the Rheumatology Department of the esteemed First Affiliated Hospital of Anhui University of Chinese Medicine. Amongst these cases, 151 patients underwent XFC treatment.A broad spectrum of indicators were encompassed in this research, includingclinical indicators [platelet count (PLT), fibrinogen (FBG), D-dimer (DD), erythrocyte sedimentation rate (ESR), high-sensitivity C-reactive protein (Hs-CRP), rheumatoid factor (RF), and anti-cyclic citrullinated peptide antibodies (CCP)]and SPP indicators [self-rating anxiety scale (SAS) score, self-rating depression scale (SDS) score, visual analog scale (VAS) score, and the medical outcomes study (MOS) 36-item short-form health survey (SF-36) assessing 8 domains: physical functioning (PF), role physical (RP), body pain (BP), general health (GH), vitality (VT), social functioning (SF), role emotional (RE), and mental health (MH)].

XFC treatment was set as “T” and unused as “F”, with indicators improved as “T” and indicators not improved as “F” after treatment. The association rule analysis between the use of XFC and the improvement of indicators was conducted using the Apriori module in IBM SPSS Modeler 18.0 software, with the support, confidence, and improvement calculated following the calculation formulas in previous research (36).

With the pharmacokinetic criteria of oral bioavailability(OB) ≥ 30% and drug-like (DL) ≥ 0.18, the active compounds and their corresponding targets of 4 TCMs were screened through the TCMSP database. Subsequently, RA-related genes were retrieved from GeneCards, OMIM, DrugBank, and TTD databases. Next, the overlapping genes between XFC active compounds and the RA targets were identified using RData software. Following this, KEGG pathway analysis was conducted on these overlapping genes to gain insights into their biological functions and interactions.

A total of 50 hospitalized RA patients from the Rheumatology Department at the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine were enrolled. Peripheral blood samples were obtained before and after XFC treatment and RA-specific PBMCs (RA-PBMCs) were isolated utilizing a density gradient method (Histopaque-1077, Sigma). All participants met the diagnostic criteria for RA, established by the American College of Rheumatology (ACR) and the European League Against Rheumatology (EULAR) in 2010 (37). Meanwhile, a total of 30 healthy volunteers, matched for age and gender, were recruited from the Physical Examination Center of the same hospital. This research strictly adhered to the ethical principles of the Helsinki Declaration. During the follow-up process, we ensured comprehensive privacy protection for patients and no disruption to their treatment plans. This study was approved by the Ethics Committee of the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine, and all participants signed the informed consent (Ethics Number: 2019AH-12).

RA-FLSs were obtained from Beijing Beina Chuanglian Biotechnology Research Institute and identified through short tandem repeat (STR) profiling. RA-FLSs were isolated and cultured following established protocols outlined by Wen et al. (31), with excess cells discarded. The remaining RA-FLSs were digested with 0.25% trypsin and resuspended in a complete Dulbecco’s Modified Eagle Medium (DMEM). To establish a co-culture system, RA-FLSs and RA-PBMCs (both 5 x10⁵ cells/well) were seeded into Transwell chambers and maintained in a cell culture incubator (37°C, 5% CO₂).

For transfection, the co-cultured RA-FLSs were collected and seeded in DMEM medium supplemented with 10% fetal bovine serum (FBS) in 24-well plates. The cells were incubated overnight in a cell culture incubator, (37°C, 5% CO₂). Subsequently, 5μL of the transfection reagent (Lipofectamine 2000) was introduced to the cells. The overexpression plasmids/siRNA wasmixed with DMEM medium, followed by incubation (48 h, 37°C) with the cells. Cells were collected and subjected to reverse transcription, and gene expression levels were quantified using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) to assess the transfection efficiency.

Cells and supernatants were collected from 50 RA patients and 30 healthy volunteers with PrimeScript™RT reagent Kit with gDNA Eraser kit following the manufacturer’s instructions. Cell RNA was extracted using the TRIzol method for the amplification reaction. PCR products were analyzed through agarose gel electrophoresis, and relative quantitative analysis was performed using the 2^-△△Ct method, with β-actin expression as a reference. The primer sequences for each detection indicator are shown in Table 1 .

Total protein was extracted from co-cultured RA-FLSs or PBMCs using a radioimmunoprecipitation assay (RIPA) lysis buffer.After protein quantification, samples were then loaded onto a gel for electrophoresis, followed bymembrane transfer. The membranes were blocked (room temperature, 2 h) in Western blot blocking solution [containing 5% skimmed milk powder or 5% bovine serum albumin (BSA) for phosphorylated proteins], and then incubated with primary antibodies against PI3K (1:500), p-PI3K (1:500), AKT (1:500), p-AKT (1:500), and RPLP2 (1:1000). After washing, a horseradish peroxidase (HRP)-conjugated secondary antibody (1:10000) was added for incubation (room temperature, 2 h), followed by 3 washeswith PBST detergent (10 min each). After that, the protein was detected using an enhanced chemiluminescence assay kit. Finally, the protein blot data were quantified and analyzed using the ImageJ software.

The supernatant of the co-cultured RA-FLSs was collected and centrifuged (1000r/min, 10 min). Next, 100 μL of supernatant was added to each well of the ELISA plate for incubation (37°C, 1.5 h). The concentration of cytokines [IL-6, IL-8, IL-4, IL-10, VEGF, and platelet-activating factor (PAF)] was detected using the ELISA kits following the protocols.

DSCR9 was synthesized using GS ™ The T7 Biotin Labeled RNA Synthesis Kit (Guangzhou Geneseed Biotech Co., Ltd., Guangzhou, China) as per the manufacturer’s instructions. The interaction between DSCR9 and RPLP2 was detected using the PureBindingTM RNA Protein pull-down Kit (Geneseed) following the manufacturer’s instructions. The RNA binding protein complex was washed, eluted, and then subjected to silver staining analysis. Afterward, Western blot was used to verify DSCR9 and RPLP2, following the same steps as before.

The co-cultured RA-FLSs were seeded onto 96-well plates and incubated with EdU diluent. Next, the cells were fixed with 4% paraformaldehyde for 30 min and then washed with phosphate-buffered saline (PBS). Cell proliferation was detected using the EdU staining reagent (RiboBio, Guangdong, China). DAPI was employed for nuclear staining, and then cells were observed under a fluorescence microscope.

A total of 30 male Sprague-Dawley (SD) rats were randomly allocated into two groups: the normal serum group (n = 10) and the XFC drug-containing serum group (n = 20). According to previous studies (36), rats in the XFC drug-containing group received a gavage of XFC suspension (0.648g/100g/d) for 7 consecutive days, while rats in the normal serum group were orally administered physiological saline (2mL/100g/d) for 7 consecutive days. One hour after the final administration, rats were anesthetized with pentobarbital sodium (50mg/kg). Rat blood was collected from the abdominal aorta into coagulation tubes and then centrifuged (3000 rpm, 10 min) for serum isolation. The serum was then incubated and inactivated in a water bath (56°C, 30 min), filtered through a 0.22 μm membrane, and stored at -80°C for subsequent analysis. All animal procedures in this study were approved by the Animal Ethics Committee of Anhui University of Traditional Chinese Medicine (AHUCM rats 2023020).

A total of 40 male SD rats (200 ± 20 g; Jiangsu Yuxi Oriental Breeding Co., Ltd) were cultured (18-22°C, 40-60% humidity). All rats were healthy and had free access to food and water. An AA model of rats was established as described in previous literature (38). Rats were randomly allocated into 4 groups (n = 10): normal (treated with daily oral gavage of 2mL/100g physiological saline), model (treated with daily oral gavage of 2mL/100g physiological saline), XFC (treated with 0.12g/100g XFC suspension), and TPL groups (treated with 0.1mg/100g TPL solution). Rats in each group were administrated by daily gavage at a dose of 2 mL/100g for 4 consecutive weeks (39). At the end of the experiment, rats were euthanized via intraperitoneal injection of pentobarbital sodium (60-100 mg/kg), and their blood was collected from the abdominal aorta for serum isolation.

Additionally, rat knee joints were excised and fixed in 4% paraformaldehyde for pathological assessment. After dehydration with gradient alcohol, clearance, and paraffin embedding, the samples were sectioned (4 μm). The sections then underwent hematoxylin and eosin (HE) staining to evaluate synovial hyperplasia, inflammatory cell infiltration, and pannus formation under light microscopy. A scoring scale ranging from 0 to 4 was adopted in this study, as reported in previous research (68). For ultrastructural analysis, synovial tissue from the rat knee joints was promptly dissected into 1 mm³ blocks, fixed, dehydrated, and prepared according to standard protocols. After ultra-thin sectioning and staining, the synovial cell ultrastructure was observed under a transmission electron microscope (TEM).

Statistical analysis and charting of experimental data were conducted using IBM SPSS Statistics 22 and GraphPad Prism 8.2 software. Quantitative data that conformed to a normal distribution were represented as mean ± standard deviation and analyzed using a t-test or one-way analysis of variance. Quantitative data that did not follow a normal distribution were represented as quartiles [M (Q25, Q75)] and analyzed using non-parametric tests. The Spearman correlation method was employed for correlation analysis. A p-value of < 0.05 indicated a statistically significant difference.

The changes in SPP and clinical laboratory indicators before and after XFC treatment were compared and analyzed. According to the results, compared with those before treatment, regarding the SPP indicators, SAS, SDS, and VAS were significantly decreased, and PF, RP, BP, GH, VT, SF, RE, and MH of RA patients were increased after treatment; clinical indicators (FBG, DD, ESR, and Hs-CRP) were remarkably decreased ( Table 2 ). These results indicated that XFC can reduce clinical inflammation and coagulation indicators in RA patients, and improve SPP of RA patients.

Through association rule analysis of clinical data of RA patients, it was found that there is a strong correlation (lift>1) between the use of XFC and the reduction of VAS, SAS, SDS, Hs-CRP, and ESR, as well as the increase of PF, BP, MH, VT, GH, and SF ( Table 3 ). The complex network diagram showed a strong correlation between XFC treatment and the improvement of inflammation indicators, coagulation indicators, and SPP ( Figure 1 ). The above results further indicated that the use of XFC can reduce clinical inflammation and coagulation indicators in RA patients, and improve their SPP.

XFC is composed of 4 TCM herbs ( Figure 2A ). Based on the pharmacokinetic characteristics of OB ≥ 30% and DL ≥ 0.18, 92 potential active ingredients and 197 targets were screened in the TCMSP database. Additionally, a total of 3577 RA-related targets were collected through the GeneCards, OMIM, DrugBank, and TTD databases. After removing the duplicate genes, 152 RA-related targets were finally obtained. Subsequently, Venn analysis was performed on 197 targets and 152 RA-related targets corresponding to XFC active ingredients using the RData software. As a result,30 overlapping genes were identified, which were potential targets for XFC treatment of RA ( Figure 2B ). KEGG pathway enrichment analysis was conducted,and the top 10 signaling pathways were selected for visual analysis based on the magnitude of the P-value. As revealed by the KEGG enrichment analysis results, PI3K/AKT was one of the key pathways for XFC treatment of RA ( Figure 2C ).

According to Spearman correlation analysis results, lncRNA DSCR9 was positively correlated with RP, BP, GH, MH, PF, SF, RE, and VT levels, and negatively correlated with SAS, SDS, and VAS levels. IL-6 was positively correlated with VAS and SAS levels, and negatively correlated with PF and RP levels. IL-8 was negatively correlated with BP and GH levels. PAF was positively correlated with RF and VAS levels. VEGF was negatively correlated with PF levels. Additionally, PLT was positively correlated with Hs-CRP, ESR, DD, and FBG levels. Similarly, FBG was positively correlated with Hs-CRP, ESR, and DD levels. DD was positively correlated with Hs-CRP and ESR levels ( Figure 3 ; Supplementary Tables S1 , S2 ). These results suggested that the inflammation in RA patients was closely related to hypercoagulability; lncRNA DSCR9, inflammation indicators, and hypercoagulability indicators were tightly related to SPP in RA patients.

Subsequently, the potential roles of lncRNA DSCR9/PI3K/AKT and inflammatory/coagulation cytokines in RA progression were investigated, A total of 30 healthy control (HC) subjects and 50 RA patients were enrolled, and no statistical difference between the two groups was found in terms of age and gender (p > 0.05, Table 4 ). Expressions of lncRNA DSCR9, PI3K, AKT, IL-6, IL-8, IL-4, IL-10, VEGF, and PAF were examined before and after XFC treatment in both groups. LncRNA DSCR9 expression was decreased, while PI3K and AKT expression levels were increased in PBMCs of RA patients. However, RT-qPCR detection results demonstrated that lncRNA DSCR9 showed an opposite expression trend in the XFC treatment group ( Figure 4A ). Western blot analysis also confirmed opposite results of PI3K and AKT levels after XFC treatment compared to those before treatment ( Figures 4B-D ). The levels of inflammation and coagulation factors were detected using ELISA. It was found that IL-6, IL-8, VEGF, and PAF levels were increased, while IL-4 and IL-10 levels were decreased in RA patients. On the contrary, XFC treatment could reverse the above results ( Figures 4E-J ).

Therefore, these results suggested that lncRNA DSCR9, PI3K, AKT, IL-6, IL-8, IL-4, IL-10, VEGF, and PAF were involved in RA progression. Among them, the expression of lncRNA DSCR9, IL-4, and IL-10 was decreased in RA-PBMCs and increased after XFC treatment. In contrast, the expression of PI3K, AKT, IL-6, IL-8, VEGF, and PAF was upregulated in RA-PBMCs, while effectively decreased after XFC treatment.

A co-culture model of RA PBMCs and RA-FLSs was established to demonstrate the effects of lncRNA DSCR9 on RA inflammation and hypercoagulability. The optimal cell ratio between RA-PBMCs and RA-FLSs was initially determined to be 2.5:1, and the optimal co-culture time was 48 h (40). According to EdU results, si-DSCR9 promoted the proliferation, while OE-DSCR9 inhibited the proliferation of co-cultured RA-FLSs ( Figures 5A, B ). ELISA results showed that IL-6, IL-8, VEGF, and PAF levels were increased, while IL-4 and IL-10 levels were decreased in the si-DSCR9 group. The OE-DSCR9 group showed opposite results, as manifested by the decreased levels of IL-6, IL-8, VEGF, and PAF ( Figures 5C, D, G, H ), and increased levels of IL-4 and IL-10 ( Figures 5E, F ). These findings indicated that knocking out DSCR9 could drive the inflammatory response and hypercoagulable state of RA, while overexpression of DSCR9 can weaken the inflammatory response and hypercoagulable state.

Subsequently, the action mechanism of lncRNA DSCR9 was further explored and its target was clarified. LncRNA DSCR9 was converted to ENSG000000230366 through the Ensemble database (https://asia.ensembl.org/index.html). Then, the target gene of lncRNA DSCR9 was predicted to be RPLP2 through the Starbase gene library (https://starbase.sysu.edu.cn/index.php). Moreover, through the IntaRNA database (https://rna.informatik.uni-freiburg.de/IntaRNA/Input.jsp?JobID=5890431&reload=true), it’s determined that lncRNA DSCR9 and RPLP2 had binding sites, with a binding energy of -12.44 kcal/mol. Mass spectrometry analysis was further utilized to identify the differentially expressed proteins pulled down by lncRNA DSCR9, and the target RPLP2 then was revealed ( Figure 6A ). Additionally, the binding of lncRNA DSCR9 to RPLP2 was verified using Western blot analysis ( Figure 6B ). Therefore, lncRNA DSCR9 may act through RPLP2.

To elucidate the mechanism of lncRNA DSCR9 regulating RA inflammation and hypercoagulability, we then detected the expression of RPLP2/PI3K/AKT in a co-culture model of RA-PBMCs and RA-FLSs. Additionally, the levels of inflammation and coagulation indicators were also evaluated after OE-RPLP2 transfection. The RT-qPCR results revealed that RPLP2 expression was increased in the si-DSCR9 group, while decreased in the OE-DSCR9 group ( Figure 7A ). As indicated by Western blot results, RPLP2, PI3K, AKT, p-PI3K, and p-AKT protein levels were increased in the si-DSCR9 group, while decreased in the OE-DSCR9 group ( Figure 7B ). After OE-DSCR9 treatment, RPLP2 expression and PI3K, AKT, p-PI3K, and p-AKT levels in the OE-RPLP2 group were decreased ( Figure 7C-E ). Similarly, OE-DSCR9 treatmentcould decrease IL-6, IL-8, IL-4, IL-10, VEGF, and PAF levels, while increase IL-4 and IL-10 levels in the OE-RPLP2 group ( Figures 7F-K ). These results indicated that lncRNA DSCR9 inhibited the PI3K/AKT pathway activation by downregulating RPLP2, thereby improving inflammation and hypercoagulability.

Whether the PI3K/AKT pathway was involved in the regulation of lncRNA DSCR9 in RA-FLSs was further explored.As PI3K is located upstream of AKT, and activating PI3K can also activate AKT, PI3K activator (RMH) was added in this study. The EdU staining results revealed that RMH can promote the proliferation of co-cultured RA-FLSs and reverse the inhibitory effect of OE-LncDSCR9 on co-cultured RA-FLS proliferation ( Figures 8A, B ). Additionally, the ELISA results showed that RMH can increase IL-6, IL-8, VEGF, and PAF levels, reduce IL-4 and IL-10 levels, and reverse the effect of OE-LncDSCR9 on these cytokines ( Figures 8C-H ). These above results indicated that the PI3K/AKT pathway activation promoted the inflammatory response and hypercoagulable state of co-cultured RA-FLSs.

The effects of different concentrations (5%, 10%, 20%, and 50%) of XFC drug-containing serum on the viability of co-cultured RA-FLS viability were detected using the CCK-8 assay at 24 h, 48 h, and 72 h, respectively. It was found that at 48 h, XFC drug-containing serum at 10% concentration had an inhibition rate of reaching 50% on co-cultured RA-FLSs. Therefore, 10% XFC-containing serum was selected as the optimal concentration, and 48 h was chosen as the optimal treatment time for the subsequent experiments (40).

XFC (10%) can inhibit the proliferation of co-cultured cells ( Figures 9A, B ), and suppress IL-6, IL-8, VEGF, and PAF levels while increasing IL-4 and IL-10 levels ( Figures 9C-H ). Additionally, XFC treatment increased lncRNA DSCR9 expression and decreased the expression levels of PI3K, AKT, p-PI3K, p-AKT, and RPLP2 ( Figures 9I-K ). Taken together, XFC (10%) exhibited similar therapeutic effects to OE-DSCR9 in inhibiting PI3K/AKT-induced inflammation and hypercoagulability.

Functional response experiments were further conducted to investigate whether XFC improved inflammation and hypercoagulability through the PI3K/AKT pathway. The EdU results showed that RMH antagonized the inhibitory effect of XFC on the proliferation of co-cultured RA-FLSs ( Figures 9L, M ). Similar results were also observed in ELISA and Western blot analysis results. In RMH-treated co-cultured RA-FLSs, IL-6, IL-8, VEGF, PAF, PI3K, AKT, p-PI3K, p-AKT, and RPLP2 expression levels were upregulated, while IL-4 and IL-10 levels were downregulated ( Figures 9N-U ). These results indicated that RMH can antagonize the therapeutic effect of XFC on the co-cultured RA-FLSs.

Finally, the rescue experiments were conducted by incorporating OE-DSCR9. The EdU results indicated that compared with the XFC (10%) + OE-NC group,the XFC (10%) + OE-DSCR9 group showed inhibited cell proliferation, and OE-DSCR9 reversed the promoting effect of RMH on cell proliferation ( Figures 9L, M ). ELISA results showed that IL-6, IL-8, VEGF, and PAF levels were remarkably reduced, while IL-4 and IL-10 levels were increased in the XFC (10%) + OE-DSCR9 group compared with those in the XFC (10%) + OE-NC group ( Figures 9N-S ). The cytokine expression trend in the XFC (10%) + RMH + OE-DSCR9 group was the same as that in the XFC (10%) + RMH group. According to Western blot analysis results, compared with those in the XFC (10%) + OE-NC group, PI3K, p-PI3K, AKT, p-AKT, and RPLP2 levels in the XFC (10%) + OE-DSCR9 group were remarkably reduced; the protein level trend in the XFC (10%) + RMH + OE-DSCR9 was the same as that in the XFC (10%) + RMH group ( Figures 9T, U ). Taken together, OE-DSCR9 can synergize with XFC to inhibit the inflammation and hypercoagulability of co-cultured RA-FLSs induced by the PI3K/AKT pathway activation. In summary, XFC can inhibit the inflammatory response and hypercoagulable state of co-cultured RA-FLSs through the LncDSCR9/RPLP2/PI3K/AKT axis.

Previous studies have confirmed that XFC can effectively improve tissue swelling and arthritis index in AA rats (41). For further evaluating whether XFC can alleviate inflammation and hypercoagulability in AA rats, RA diagnostic indicators (RF and CRP), inflammatory factors (IL-6and IL-10), and coagulation indicators (FBG and DD) were detected. The pathological and ultrastructural changes in AA rats were also observed. The results showed that compared with those in the normal group, the serum levels of RF, CRP, IL-6, FBG, and DD in AA rats were notably increased, while IL-10 level was remarkably reduced ( Figures 10A-F ). This indicated that the inflammation and hypercoagulability were promoted and anti-inflammatory ability was inhibited in AA rats. After XFC and TPL treatment, the above indicators were effectively improved ( Figures 10A-F ).

According to HE staining results, AA rats showed synovial tissue proliferation, accompanied by extensive infiltration of inflammatory cells, narrow joint space, uneven joint surface, and pannus formation. All these were consistent with the basic pathological characteristics of human RA ( Figure 10G ). Safranin-O-Fast Green staining results revealed disrupted knee joint structure, damaged bone integrity, synovial tissue-invaded cartilage, degraded and damaged red cartilage matrix, and rough surface of the joint cartilage in AA rats ( Figure 10G ). After XFC and TPL treatment, the surface of the knee joint in rats was relatively intact, and the degradation of the cartilage matrix was reduced (p < 0.05, Figure 10J ). This indicated that XFC efficiently inhibited synovial hyperplasia, inflammatory cell infiltration, and cartilage destruction in AA rats.

As indicated by immunohistochemical staining results, there were brownish-yellow granular substances (PI3K and AKT) in the cells ( Figure 10H ). The treatment with XFC and TPL reduced the expression of PI3K and AKT, suggesting that XFC may exert anti-inflammatory and anticoagulant effects through the PI3K/AKT pathway. TEM observed that the number of organelles in synovial cells of the model group rats was significantly reduced, accompanied by mitochondrial vacuolar degeneration, and mitochondrial cristae rupture or even disappearance. There was also significant expansion and damage to the rough endoplasmic reticulum in the model rats. After XFC and TPL treatments, the nuclear membrane boundaries of synovial cells were clear, with mild mitochondrial swelling, clear mitochondrial cristae, and normal morphology of the rough endoplasmic reticulum ( Figure 10I ). The above results indicated that XFC can improve the inflammatory response and hypercoagulable state of AA rats.