Translate this page into:
Identification of key pathways and biomarkers in rheumatoid arthritis synovial tissue through comprehensive transcriptomic integration
⁎Corresponding author: Jinzhou Luo. l025690@163.com
⁎⁎Corresponding author: Mao-Lin He. hemaolin@stu.gxmu.edu.cn
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by synovitis and joint destruction. Although our understanding of its pathogenesis has deepened, the molecular mechanisms underlying changes in synovial tissue remain incompletely elucidated.
We integrated transcriptomic datasets from the GEO database, including gene expression microarray datasets and RNA-seq datasets (GSE1919, GSE12021, GSE55235, GSE55457, GSE77298, GSE89408), to identify differentially expressed genes (DEGs) in RA synovial tissue. Functional enrichment (GO, KEGG) and gene set enrichment analysis (GSEA) were performed to explore key pathways. CIBERSORT was used to assess immune cell infiltration. We applied a comprehensive machine learning approach using 113 algorithms to screen for core diagnostic genes, which were subsequently validated across multiple datasets.
In the training set (GSE89408), we identified 9204 DEGs that were significantly enriched in immune-related processes (leukocyte migration, cytokine activity) and pathways (cytokine-cytokine receptor interaction, chemokine signaling). GSEA confirmed the activation of these pathways (NES >1, FDR <0.001). Immune infiltration analysis showed a significant increase in plasma cells in RA synovium (with no plasma cells in the control group). Machine learning identified 9 core genes (AIM2, AKR1B10, CXCL10, CXCL13, IGLC1, IL2RG, LRRC15, SDC1, and IGHG1), which demonstrated robust diagnostic performance.
Cytokine-cytokine receptor interactions, chemokine signaling pathways, and plasma cell infiltration may play critical roles in synovial pathogenesis of rheumatoid arthritis. Among the 9 core genes identified, 8 have been experimentally validated for their functions in RA, while the unverified IGHG1 may serve as an important biomarker for the pathogenesis of RA synovium.
Keywords
Rheumatoid arthritis (RA)
Synovial tissue
113 machine learning
IGHG1
Biomarkers
1 Introduction
Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by synovial inflammation, proliferation of synovial cells, and subsequent destruction of cartilage and bone. The pathogenesis of RA involves complex interactions among immune cells and persistent inflammation.1–4 According to epidemiological data from 2021, the global age-standardized incidence rate (ASIR) of RA among adolescents and young adults is 9.46 per 100,000 (95 % UI: 5.92 to 13.54), with an age-standardized prevalence (ASPR) of 104.35 per 100,000 (77.44–137.84), and an age-standardized mortality rate (ASMR) of 0.016 per 100,000 (0.013–0.019). Alarmingly, both the incidence and prevalence of RA are on the rise worldwide.5
Multiple factors contribute to the etiology of RA, including abnormalities in the immune system, alterations in cellular behavior, epigenetic regulation, angiogenesis, and the interactions between various signaling pathways. Together, these mechanisms lead to the chronic inflammatory state and joint destruction commonly observed in RA.6–11 Patients frequently present with generalized fatigue, joint swelling and tenderness, as well as morning stiffness in the early stages of the disease. If not treated promptly and effectively, patients may develop severe systemic manifestations during disease progression, including pleural effusions, pulmonary nodules, and interstitial lung disease.12,13
Despite these insights, the pathogenesis of local synovial tissue in RA patients remains incompletely understood. This study aims to integrate transcriptome datasets from RA synovial tissues to further investigate the infiltration characteristics of immune cells and to identify key functional pathways and core genes. The findings of this study will provide a theoretical foundation for elucidating the pathological mechanisms underlying RA synovial tissue and will contribute to a deeper understanding of the disease's pathogenesis.
2 Materials and methods
2.1 Data download and processing
The technical roadmap for this study is illustrated in Fig. 1. We retrieved transcriptome datasets related to RA from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) using the search keyword “Rheumatoid Arthritis.” The screening criteria included “Expression profiling by array,” “Expression profiling by high throughput sequencing,” and “Homo sapiens.” Additionally, the selected datasets had to contain both healthy controls and RA groups, with a minimum of 5 samples in each group. Ultimately, six datasets were included: GSE1919, GSE12021, GSE55235, GSE55457, GSE77298, and GSE89408. Detailed information about the datasets is provided in Table 1. Notably, GSE89408 consists of next-generation sequencing (high-throughput sequencing) data, while the other five datasets are derived from first-generation sequencing. For data analysis, GSE89408 was utilized as the training set, and the other microarray datasets served as validation sets.
| Dataset | Platform ID | Normal Samples | RA Samples | Species |
| GSE1919 | GPL91 | 5 | 5 | Human |
| GSE12021 | GPL96 | 9 | 12 | Human |
| GSE55235 | GPL96 | 10 | 10 | Human |
| GSE55457 | GPL96 | 10 | 13 | Human |
| GSE77298 | GPL570 | 7 | 16 | Human |
| GSE89408 | GPL570 | 23 | 150 | Human |
2.2 Differential gene analysis
Differentially expressed gene (DEG) analysis was conducted on all datasets using the “limma” R package, with screening criteria set at a corrected p-value <0.05 and |log fold change (FC)| ≥ 1. The “ggplot2″ and “pheatmap” R packages were employed to create volcano plots and heat maps to visualize the expression of differentially expressed genes.
2.3 Enrichment analysis of GO and KEGG
Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed using the “clusterProfiler” R package. Statistical significance was determined at a corrected p-value <0.05, and the enrichment of differential genes in cell components, biological processes, molecular functions, and pathways was analyzed.
2.4 GSEA analysis
Gene Set Enrichment Analysis (GSEA) was performed using the “clusterProfiler” R package, applying a filtering criterion of corrected p-values <0.05. Pathways with a normalized enrichment score (NES) greater than 1 were considered significantly activated, while those with NES less than 1 were deemed significantly inhibited. GSEA analysis was utilized to identify pathways activated or inhibited during the development of the disease.14
2.5 Immune cell infiltration analysis
The deconvolution algorithm and the “CIBERSORT” R package were employed to infer the infiltration of various immune cells in the synovial tissues of normal individuals and rheumatoid arthritis patients. Bar charts were generated to illustrate the composition of immune cells in RA samples compared to normal samples, and violin plots were employed to visually depict differences in the expression levels of different immune cells between RA and normal samples.
2.6 Deep screening of core genes using machine learning
Based on the differentially expressed genes identified in the training set, we applied a total of 113 machine learning algorithms for an in-depth analysis across all datasets. These algorithms were arranged and combined using 12 basic algorithms, including LASSO, SVM, Random Forest (RF), glmBoost, Partial Least Squares Regression (plsRglm), Stepglm, Elastic Net (Enet), Linear Discriminant Analysis (LDA), Gradient Boosting Machine (GBM), Ridge Regression, XGBoost, and Naive Bayes. The R packages utilized included: randomForestSRC, glmnet, plsRglm, gbm, caret, mboost, e1071, MASS, xgboost, and stats. The area under the receiver operating characteristic (ROC) curve (AUC) for the training set and five validation sets was calculated for each of the 113 machine learning algorithms, with model results displayed as heat maps. Genes screened by the algorithm yielding the largest average AUC value were designated as core genes. Finally, the ROC curve and mRNA expression of core genes across all datasets were plotted.
3 Results
3.1 Identification of differentially expressed genes
The differential gene expression across all datasets is illustrated in Fig. 2A and B. A total of 9204 DEGs were identified in the training set, of which 4015 were up-regulated and 5189 were down-regulated. In GSE1919, 459 DEGs were identified, with 234 up-regulated and 225 down-regulated. GSE12021 revealed 253 DEGs, including 140 up-regulated and 113 down-regulated genes. In GSE55235, a total of 1108 DEGs were identified, with 619 up-regulated and 489 down-regulated genes. GSE55457 identified 341 DEGs, comprised of 210 up-regulated and 131 down-regulated genes. Lastly, GSE77298 included 468 DEGs, with 260 up-regulated and 208 down-regulated genes.
3.2 GO and KEGG enrichment analysis of aggregated immune responses
The GO enrichment analysis circle diagram shows the entries of these differentially expressed genes that were significantly enriched in biological process (BP), cell composition (CC) and molecular function (MF) (Fig. 3A). However, the circle plot could not show all the entries with significant differences, so we compiled the simultaneously enriched entries from all datasets into Table 2. These genes were mainly involved in several key biological processes, such as the migration, regulation and chemical orientation of various types of leukocytes (including lymphocytes, monocytes and neutrophils), cell-cell interactions, and various immune responses (leukocyte-mediated immune response, adaptive immune response and regulatory immune effector processes). In terms of cellular composition, these genes are involved in the outer surface of the plasma membrane; In terms of molecular function, it focuses on cytokine activity and CXCR chemokine receptor binding. KEGG enrichment analysis showed that the pathways that were common to all the data sets were the cytokine-cytokine receptor interaction pathway and the chemokine signaling pathway (Fig. 3B).
| ONTOLOGY | ID | Description |
| BP | GO:0022407 | regulation of cell-cell adhesion |
| BP | GO:0050900 | leukocyte migration |
| BP | GO:0002443 | leukocyte mediated immunity |
| BP | GO:0002449 | lymphocyte mediated immunity |
| BP | GO:0030595 | leukocyte chemotaxis |
| BP | GO:0060326 | cell chemotaxis |
| BP | GO:0001909 | leukocyte mediated cytotoxicity |
| BP | GO:0071674 | mononuclear cell migration |
| BP | GO:1990266 | neutrophil migration |
| BP | GO:0097529 | myeloid leukocyte migration |
| BP | GO:0002460 | adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily domains |
| BP | GO:0030593 | neutrophil chemotaxis |
| BP | GO:0097530 | granulocyte migration |
| BP | GO:0072676 | lymphocyte migration |
| BP | GO:0071621 | granulocyte chemotaxis |
| BP | GO:0002697 | regulation of immune effector process |
| BP | GO:0002685 | regulation of leukocyte migration |
| BP | GO:1990868 | response to chemokine |
| BP | GO:1990869 | cellular response to chemokine |
| BP | GO:0034612 | response to tumor necrosis factor |
| BP | GO:0071675 | regulation of mononuclear cell migration |
| BP | GO:0050853 | B cell receptor signaling pathway |
| BP | GO:0006816 | calcium ion transport |
| CC | GO:0009897 | external side of plasma membrane |
| MF | GO:0005125 | cytokine activity |
| MF | GO:0045236 | CXCR chemokine receptor binding |
3.3 GSEA analysis showed that the significantly activated pathways were highly consistent with KEGG
A gene-set enrichment analysis of the classic 180 KEGG pathways showed that only two pathways were co-activated or inhibited in all data sets: the cytokine-cytokine receptor interaction pathway and the chemokine signaling pathway (Fig. 4A and B). Both pathways showed significant activation signatures in all data sets, with NES greater than 1 and an adjusted p value of less than 0.001.
More notably, the previous KEGG analysis results also showed that differentially expressed genes were significantly enriched in these two pathways. This high concordance further emphasizes the importance of cytokine-cytokine receptor interaction pathway and chemokine signaling pathway in the development of RA.
3.4 Immune infiltration analysis revealed significant infiltration of plasma cells in RA synovial tissue
The histogram illustrates the ratio of 22 immune cell types infiltrating normal versus RA synovial tissue (Fig. 5A). Violin plots provide a visual representation of the significant differences in immune cell infiltration between the normal and RA groups. While variations in immune cell infiltrations were observed across different datasets, a consistent finding emerged: the proportion of plasma cells in the normal group was nearly absent, whereas a significant infiltration of plasma cells was observed in the RA group (Fig. 5B). This suggests a marked increase in plasma cell populations within the synovial tissue of RA patients.
3.5 Screening of nine core genes using 113 machine learning algorithms
Based on 113 machine learning algorithms, we conducted an in-depth analysis of all datasets to identify the best predictive model. Multiple models showed excellent performance, with AUC values exceeding 0.95, demonstrating significant predictive power (Fig. 6A). In particular, the average AUC of the plsRglm model reached an impressive 0.968, so we selected this model as the best model and screened nine core genes: AIM2, AKR1B10, CXCL10, CXCL13, IGHG1, IGLC1, IL2RG, LRRC15, and SDC1. We then plotted ROC curves and their mRNA expression profiles for the nine core genes (Fig. 6B). The results showed that the AUC values of these core genes were more than 0.70 in the training set and all validation sets. Among them, AIM2, CXCL10, CXCL13, IL2RG and SDC1 had AUCs greater than 0.8 in all datasets. The mRNA expression levels of these nine genes were shown by boxplots (Fig. 7A–F), and the results showed that they were able to significantly distinguish the normal group from the RA group (all p values < 0.05). AKR1B10 was significantly down-regulated in the RA group compared with the normal group, while the remaining core genes were significantly up-regulated in the RA group.
4 Discussion
In this study, we conducted a comprehensive transcriptome analysis of RA synovial tissue, integrating both first-generation and next-generation sequencing technologies to explore the molecular characteristics of RA. Based on GO analysis, we found that the migration, regulation, and chemotaxis of various leukocytes, cell-cell interactions, and multiple immune responses are closely related to the occurrence and progression of RA. Moreover, KEGG and GSEA consistently demonstrated significant activation of the cytokine-cytokine receptor interaction pathway and chemokine signaling pathway during the development of RA. Immune infiltration analysis revealed significant infiltration of plasma cells within RA synovial tissues. By applying 113 machine learning algorithms, we successfully identified nine key core genes (AIM2, AKR1B10, CXCL10, CXCL13, IGHG1, IGLC1, IL2RG, LRRC15, and SDC1), which exhibited strong diagnostic performance in distinguishing normal controls from RA patients. Thus, we propose that these nine core genes may serve as important biomarkers influencing RA synovial tissue.
This study has notable advantages over previous investigations. First, regarding sequencing technology, we integrated both first-generation and second-generation sequencing data of RA synovial tissue, while most prior studies relied solely on first-generation sequencing (microarray data), which lacked comprehensiveness. Second, our study incorporated a large sample size, totaling 270 samples— including 64 healthy controls and 206 RA samples—compared to smaller sample sets used in previous studies, some of which lacked appropriate normal or disease samples. We discarded datasets with sample sizes less than five to mitigate potential biases.15–19 Third, our analysis highlights commonalities among transcriptomic studies conducted by different researchers in varied experimental settings and time points. This is crucial, as differences in clinical sampling techniques may lead to disparate analysis results. By integrating diverse datasets, we aimed to draw conclusions with broader applicability. Finally, our unique approach utilizes a combination of 113 algorithms, whereas previous studies typically employed only 2–3 algorithms for intersection. Given that each algorithm has its strengths and weaknesses, using a more extensive combination can help avoid overlooking potentially important genes.20–23
Our KEGG and GSEA analyses underscore the critical roles of the cytokine-cytokine receptor interaction pathway and chemokine signaling pathway in RA. Although several prior studies have indicated that various pathways may be significantly activated or inhibited in RA pathogenesis, we found that certain pathways did not demonstrate substantial changes across these studies (p < 0.05). Furthermore, discrepancies exist among researchers regarding the degree of immune cell infiltration in RA synovial tissue. In contrast, our findings revealed a consistently higher proportion of plasma cells in RA synovial tissue when compared to normal controls across all datasets, with normal synovial tissue almost entirely devoid of plasma cells. This observation suggests that plasma cells may play a pivotal role in the pathogenesis of RA, warranting further investigation in future studies.
We observed significantly increased expression levels of AIM2 in the synovial tissue and fibroblast-like cells of RA patients compared to healthy controls. Previous studies have indicated that AIM2 knockdown impairs the proliferation, migration, and invasion of fibroblast-like synoviocytes.24,25 AKR1B10 functions as an inflammation regulator, stimulating its own expression.26 In rats, the mRNA expression level of AKR1B10 in synovial tissue was markedly lower than that in controls.27 Moreover, low expression of AKR1B10 in gastric cancer tissues has been associated with M2 macrophage polarization, promoting disease progression.28 Thus, it is plausible that AKR1B10 may also influence macrophage polarization in synovial tissues. Additionally, we confirmed the significantly higher expression of CXCL13 and CXCL10 in RA synovium, with JAK inhibitor tofacitinib effectively inhibiting JAK1-STAT signaling and reducing mRNA expression levels of these chemokines.29,30
L2RG is integral to the development, differentiation, and function of T cells, B cells, and NK cells, with its deletion leading to severe immune system defects.31Compared to healthy individuals, IL2RG expression was significantly up-regulated.32 Studies indicate that overexpression of LRRC15 correlates with enhanced proliferation, migration, invasion, and angiogenesis of RA synovial fibroblasts. Inhibiting LRRC15 expression can significantly curtail synovial proliferation and bone invasion, thereby potentially mitigating joint destruction.33,34 Furthermore, serum levels of SDC-1 were significantly elevated in RA patients compared to normal controls.35 However, the role of SDC-1 remains controversial, as some research suggests its loss may exacerbate the inflammatory features of RA, indicating the need for further experimental validation.36 In terms of gene expression, IGLC1 mRNA levels were notably increased in RA patients relative to healthy samples, a trend corroborated at the protein level.37 Although no study has directly investigated IGHG1 expression in RA synovium, evidence suggests it may promote disease development and metastasis.38–40
In summary, our study makes three significant contributions. First, in terms of pathway analysis, we identified two activated signaling pathways. Second, we highlighted the significant infiltration of plasma cells in RA synovium regarding immune cell infiltration. Finally, we identified nine key core genes in RA synovial tissue, eight of which have been experimentally validated.
Nevertheless, some limitations remain in this study. The activation of the identified pathways and the novel gene IGHG1 have yet to be validated through in vitro and animal models. Therefore, we anticipate that our findings can provide new insights and directions for future research on RA synovial tissue.
5 Conclusions
In this study, KEGG and GSEA analyses identified significant up-regulation of cytokine-cytokine receptor interaction pathway and chemokine signaling pathway in RA synovium. In addition, immune infiltration analysis revealed significant infiltration of plasma cells. Nine core genes (AIM2, AKR1B10, CXCL10, CXCL13, IGHG1, IGLC1, IL2RG, LRRC15, SDC1) were identified and validated using 113 machine learning algorithms, providing new insights into the understanding of the molecular mechanisms of RA.
Guardian/patient's consent
The data for this study were sourced from the GEO database, which is a public repository for gene expression data. The data available in the GEO database are comprised of studies that have been submitted by researchers who have obtained the necessary approvals from their respective institutional review boards and informed consent from the participating subjects. Therefore, this study does not require additional proof of informed consent from the subjects.
Author statement
Qifan Chen: Conceptualization; Methodology; Formal analysis; Writing - Original Draft.
Chu-Song Zhou: Conceptualization; Methodology.
Hanhua Wu: Software; Validation.
Bufan Li: Investigation; Data Curation.
Yu-Nan Man: Data Curation; Visualization.
Mao-Lin He: Writing - Review & Editing.
Jinzhou Luo: Writing - Review & Editing.
Ethical statement
The data for this study were sourced from the GEO database, which is a public repository for gene expression data. The GEO database ensures that all data submissions comply with relevant guidelines and regulations for the collection and sharing of biological data. Therefore, this study does not require additional ethical or moral statements.
Funding
This work was supported by the National Natural Science Foundation of China under Grant [grant number 82160536].
References
- NR1D1 modulates synovial inflammation and bone destruction in rheumatoid arthritis. Cell Death Dis. 2020;11(2)
- [Google Scholar]
- Age-associated B cells contribute to the pathogenesis of rheumatoid arthritis by inducing activation of fibroblast-like synoviocytes via TNF-alpha-mediated ERK1/2 and JAK-STAT1 pathways. Ann Rheum Dis. 2022;81(11):1504-1514.
- [Google Scholar]
- Arachidonic acid-regulated calcium signaling in T cells from patients with rheumatoid arthritis promotes synovial inflammation. Nat Commun. 2021;12(1):907.
- [Google Scholar]
- Macrophage-derived mir-100-5p orchestrates synovial proliferation and inflammation in rheumatoid arthritis through mTOR signaling. J Nanobiotechnol. 2024;22(1):197.
- [Google Scholar]
- Age-standardized incidence, prevalence, and mortality rates of autoimmune diseases in adolescents and young adults (15–39 years): an analysis based on the global burden of disease study 2021. BMC Public Health. 2024;24(1)
- [Google Scholar]
- Rheumatoid Arthritis - Common origins, divergent mechanisms. N Engl J Med. 2023;388(6):529-542.
- [Google Scholar]
- Persistent inflammatory and non-inflammatory mechanisms in refractory rheumatoid arthritis. Nat Rev Rheumatol. 2021;17(1):17-33.
- [Google Scholar]
- RelA/MicroRNA-30a/NLRP3 signal axis is involved in rheumatoid arthritis via regulating NLRP3 inflammasome in macrophages. Cell Death Dis. 2021;12(11):1060.
- [Google Scholar]
- Pathogenesis, clinical features, and treatment strategy for rheumatoid arthritis-associated interstitial lung disease. Autoimmun Rev. 2022;21(5)
- [Google Scholar]
- Rheumatoid arthritis: Extra-articular manifestations and comorbidities. Autoimmun Rev. 2021;20(4)
- [Google Scholar]
- Identification of potential ferroptosis-associated biomarkers in rheumatoid arthritis. Front Immunol. 2023;14
- [Google Scholar]
- Bioinformatics analysis and identification of genes and molecular pathways involved in synovial inflammation in rheumatoid arthritis. Med Sci Monit. 2019;25:2246-2256.
- [Google Scholar]
- Identification of hub genes in rheumatoid arthritis through an integrated bioinformatics approach. J Orthop Surg Res. 2021;16(1):458.
- [Google Scholar]
- Identification of key genes in rheumatoid arthritis and osteoarthritis based on bioinformatics analysis. Medicine (Baltim). 2018;97(22)
- [Google Scholar]
- Identification of potential ferroptosis key genes and immune infiltration in rheumatoid arthritis by integrated bioinformatics analysis. Heliyon. 2023;9(11)
- [Google Scholar]
- Identification of PSMB9 and CXCL13 as immune-related diagnostic markers for rheumatoid arthritis by machine learning. Curr Pharm Des. 2022;28(34):2842-2854.
- [Google Scholar]
- Bioinformatics analysis of rheumatoid arthritis tissues identifies genes and potential drugs that are expressed specifically. Sci Rep. 2023;13(1):4508.
- [Google Scholar]
- Analysis and experimental validation of rheumatoid arthritis innate immunity gene CYFIP2 and pan-cancer. Front Immunol. 2022;13
- [Google Scholar]
- Machine learning to identify immune-related biomarkers of rheumatoid arthritis based on WGCNA network. Clin Rheumatol. 2022;41(4):1057-1068.
- [Google Scholar]
- Unveiling the link between lactate metabolism and rheumatoid arthritis through integration of bioinformatics and machine learning. Sci Rep. 2024;14(1):9166.
- [Google Scholar]
- Expression of AIM2 in rheumatoid arthritis and its role on fibroblast-like synoviocytes. Mediat Inflamm. 2020;2020
- [Google Scholar]
- Myricitrin inhibits fibroblast-like synoviocyte-mediated rheumatoid synovial inflammation and joint destruction by targeting AIM2. Front Pharmacol. 2022;13
- [Google Scholar]
- Aldose reductase: new insights for an old enzyme. Biomol Concepts. 2011;2(1-2):103-114.
- [Google Scholar]
- Exploring the therapeutic potential of "Tianyu" medicine pair in rheumatoid arthritis: an integrated study combining LC-MS/MS, bioinformatics, network pharmacology, and experimental validation. Front Med (Lausanne). 2024;11
- [Google Scholar]
- AKR1B10 regulates M2 macrophage polarization to promote the malignant phenotype of gastric cancer. Biosci Rep. 2023;43(10)
- [Google Scholar]
- The JAK inhibitor tofacitinib suppresses synovial JAK1-STAT signalling in rheumatoid arthritis. Ann Rheum Dis. 2015;74(6):1311-1316.
- [Google Scholar]
- The bromodomain protein inhibitor I-BET151 suppresses expression of inflammatory genes and matrix degrading enzymes in rheumatoid arthritis synovial fibroblasts. Ann Rheum Dis. 2016;75(2):422-429.
- [Google Scholar]
- Generation of inactivated IL2RG and RAG1 monkeys with severe combined immunodeficiency using base editing. Signal Transduct Targeted Ther. 2023;8(1):327.
- [Google Scholar]
- Screening and identification of potential hub genes and immune cell infiltration in the synovial tissue of rheumatoid arthritis by bioinformatic approach. Heliyon. 2023;9(1)
- [Google Scholar]
- RUNX1 ameliorates rheumatoid arthritis progression through epigenetic inhibition of LRRC15. Mol Cells. 2023;46(4):231-244.
- [Google Scholar]
- Identification and validation of hub genes for predicting treatment targets and immune landscape in rheumatoid arthritis. BioMed Res Int. 2022;2022
- [Google Scholar]
- Serum levels of free light chains and syndecan-1 in patients with rheumatoid arthritis and systemic lupus erythematosus. Rheumatology (Oxford). 2025;64(5):2422-2431.
- [Google Scholar]
- Lack of Syndecan-1 promotes the pathogenesis of experimental rheumatoid arthritis. Immunogenetics. 2024;76(3):145-154.
- [Google Scholar]
- Identification of CD8(+) T cell-related biomarkers and immune infiltration characteristic of rheumatoid arthritis. Aging (Albany NY). 2024;16(2):1399-1413.
- [Google Scholar]
- IGHG1 upregulation promoted gastric cancer malignancy via AKT/GSK-3beta/beta-Catenin pathway. Cancer Cell Int. 2021;21(1):397.
- [Google Scholar]
- Silencing of IGHG1 reverses the resistance of pancreatic cancer to multidrug chemotherapy by modulating autophagy. Environ Toxicol. 2023;38(8):1835-1845.
- [Google Scholar]
- IGHG1 promotes motility likely through epithelial-mesenchymal transition in ovarian cancer. Chin J Cancer Res. 2018;30(2):282-290.
- [Google Scholar]