Machine learning identification of LRRC15 and MICB as immunodiagnostic markers for rheumatoid arthritis

doi:10.12307/2025.361

Abstract

Abstract: BACKGROUND: Rheumatoid arthritis is a chronic autoimmune disease. Early diagnosis is crucial for preventing disease progression and for effective treatment. Therefore, it is of significance to investigate the diagnostic characteristics and immune cell infiltration of rheumatoid arthritis.
OBJECTIVE: Based on the Gene Expression Omnibus (GEO) database, to screen potential diagnostic markers of rheumatoid arthritis using machine learning algorithms and to investigate the relationship between the diagnostic characteristics of rheumatoid arthritis and immune cell infiltration in this pathology.
METHODS: The gene expression datasets of synovial tissues related to rheumatoid arthritis were obtained from the GEO database. The data sets were merged using a batch effect removal method. Differential expression analysis and functional correlation analysis of genes were performed using R software. Bioinformatics analysis and three machine learning algorithms were used for the extraction of disease signature genes, and key genes related to rheumatoid arthritis were screened. Furthermore, we analyzed immune cell infiltration on all differentially expressed genes to examine the inflammatory state of rheumatoid arthritis and investigate the correlation between their diagnostic characteristics and infiltrating immune cells.
RESULTS AND CONCLUSION: In both rheumatoid arthritis and normal synovial tissues, we identified 179 differentially expressed genes, with 124 genes up-regulated and 55 genes down-regulated. Enrichment analysis revealed a significant correlation between rheumatoid arthritis and immune response. Three machine learning algorithms identified LRRC15 and MICB as potential biomarkers of rheumatoid arthritis. LRRC15 (area under the curve=0.964, 95% confidence interval: 0.924-0.992) and MICB (area under the curve=0.961, 95% confidence interval: 0.923-0.990) demonstrated strong diagnostic performance on the validation dataset. The infiltration of 13 types of immune cells was altered, with macrophages being the most affected. In rheumatoid arthritis, the majority of proinflammatory pathways in immune cell function were activated. Immunocorrelation analysis revealed that LRRC15 and MICB had the strongest correlation with M1 macrophages. To conclude, this study identified LRRC15 and MICB as potential diagnostic markers for rheumatoid arthritis, with strong diagnostic performance and significant correlation with immune cell infiltration. Machine learning and bioinformatics analysis deepened the understanding of immune infiltration in rheumatoid arthritis and provided new ideas for the diagnosis and treatment of rheumatoid arthritis.

中国组织工程研究杂志出版内容重点：组织构建；骨细胞；软骨细胞；细胞培养；成纤维细胞；血管内皮细胞；骨质疏松；组织工程

Key words: rheumatoid arthritis, machine learning, immune infiltration, diagnostic marker, differentially expressed gene

CLC Number:

Tian Yanhu, Huang Xinan, Guo Tongtong, Rusitanmu·Ahetanmu, Luo Jiangmiao, Xiao Yao, Wang Chao, Wang Weishan . Machine learning identification of LRRC15 and MICB as immunodiagnostic markers for rheumatoid arthritis[J]. Chinese Journal of Tissue Engineering Research, 2025, 29(11): 2411-2420.

Figures/Tables 10

在显著差异。上调最显著的前10个基因分别是MMP1、LOC101928620、CXCL13、IGHM、IGJ、IGLV1-44、IGLC1、CXCL9、IGKC、LRRC15，下调最显著的前10个基因分别是ABCA8、ZBTB16、MAOA、PLIN1、FOSB、RPS4Y1、APOD、FABP4、ADIPOQ、ADH1B。 2.2 富集分析此次研究进行了功能富集分析，以更深入地了解差异表达基因的生物学功能。GO富集分析结果显示，差异表达基因的生物学过程主要富集在白细胞-细胞黏附、细胞-细胞黏附的调节、免疫反应的激活、单核细胞分化、细胞黏附的正向调节在细胞成分方面，差异表达基因主要与质膜外侧、含胶原蛋白的细胞外基质、内细胞载体相关(图2A，B)。在分子功能方面，差异表达基因主要富集在G 蛋白偶联受体结合、受体配体活性、糖胺聚糖结合方面。 KEGG富集分析显示，差异表达基因在免疫相关通路中显著富集，如细胞因子-细胞因子受体相互作用、趋化因子信号通路和病毒蛋白与细胞因子及细胞因子受体的相互作用、细胞黏附分子(图2C)。DO富集分析显示，差异表达基因主要与动脉硬化、细菌性传染病、肝炎和原发性免疫缺陷相关(图2D)。对合并后的每个滑膜组织基因进行GSEA富集分析，GOBP主要富集在免疫反应的激活、适应性免疫应答、抗原受体介导的信号通路(图2E)，B细胞受体信号通路、免疫反应调节信号通路。KEGG主要富集在同种异体移植排斥反应、抗原加工和呈递、甲状腺自身免疫性疾病、细胞黏附分子和趋化因子信号通路(图2F)。 2.3 基于机器学习的诊断标志物筛选研究使用了3种机器学习算法来识别类风湿关节炎的特征基因，对差异表达基因的表达矩阵进行LASSO回归分析(图3A)，模型的二项偏差最小，lambda值为16，选择的预测基因为LRRC15、IGHG1、PSMB9、MICB、AIM1、EDNRB、RRM2、TNFAIP8、SFRP1、QPCT、JUNB、GZMB、FKBP5、GHR、GPR64、CYP4B1；对差异表达基因的表达矩阵进行SVM-RFE算法(图3B)，根据算法的迭代过程，具有最小均方根误差(RMSE)的变量"

为LRRC15、VOPP1、MICB、CXCL13；利用随机森林结合特征选择，以相对相关性由高到低的顺序，挑选重要性评分大于2的基因(图3C)，相关的基因为LOC101928620、LRRC15、MICB、VOPP1、IGHG1，利用维恩图比较差异表达基因和关键模块基因的重叠区域(图3D)，可以识别出2个重叠基因区域，这3种机器学习算法确定了LRRC15、MICB为潜在的诊断标志物。 2.4 诊断模型的评估及列线图的构建在合并后的训练集和验证集中2个潜在标记的基因具有显著的差异表达，并且验证了2个潜在诊断标志物的诊断能力。如图4所示，在训练集中LRRC15曲线下面积为0.964，MICB曲线下面积为0.961(图4A)，2种基因合并后诊断模型曲线下面积为0.985；在验证集GSE89408中，用小提琴图展示了LRRC15及MICB的基因表达差异(图4C，D)，并且验证2种潜在的诊断标记物的诊断性能，LRRC15曲线下面积为0.759，MICB曲线下面积为0.863(图4E)，2种基因合并后诊断模型曲线下面积为0.878，结果表明在验证集中MICB具有良好的诊断能力，但LRRC15在验证集中诊断能力较差，两者合并后的诊断模型仍具有较高的诊断性能。基于2个筛选的潜在诊断标记物构建列线图来预测类风湿关节炎的发生(图5A)，标定曲线验证了模型的准确性(图5B)，决策曲线揭示了列线图优异的决策能力(图5C)。这表明构建的列线图具有良好的预测精度和可靠性。 2.5 免疫细胞浸润分析此次研究分析了类风湿关节炎和正常滑膜组织中免疫细胞的浸润情况，用直方图表示每个样本免疫细胞的组成(图6A)，结果表明，类风湿关节炎的滑膜组织中M2巨噬细胞、CD8+ T细胞、静息的肥大细胞、Tfh细胞是主要的浸润性免疫细胞。然后评估类风湿关节炎滑膜组织中免疫细胞的相关性(图6B)，例如M2巨噬细胞与浆细胞、单核细胞集中性粒细胞呈负相关，中性粒细胞与静息自然杀伤细胞、单核细胞和幼稚CD4+ T细胞呈正相关，Tfh细胞与记忆B细胞、CD8+ T细胞、幼稚的B细胞和M1巨噬细胞呈正相关。"

References

[1] MCINNES IB, SCHETT G. The pathogenesis of rheumatoid arthritis. N Engl J Med. 2011;365(23):2205-2219.
[2] DI MATTEO A, BATHON JM, EMERY P. Rheumatoid arthritis. Lancet. 2023; 402(10416):2019-2033.
[3] LEE DM, WEINBLATT ME. Rheumatoid arthritis. Lancet. 2001;358(9285):903-911.
[4] SILMAN AJ, PEARSON JE. Epidemiology and genetics of rheumatoid arthritis. Arthritis Res. 2002;4 Suppl 3(Suppl 3):S265-S272.
[5] ALETAHA D, NEOGI T, SILMAN AJ, et al. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62(9):2569-2581.
[6] BARTOK B, FIRESTEIN GS. Fibroblast-like synoviocytes: key effector cells in rheumatoid arthritis. Immunol Rev. 2010; 233(1):233-255.
[7] WEYAND CM, GORONZY JJ. The immunology of rheumatoid arthritis. Nat Immunol. 2021; 22(1):10-18.
[8] MÜLLER-LADNER U, OSPELT C, GAY S, et al. Cells of the synovium in rheumatoid arthritis. Synovial fibroblasts. Arthritis Res Ther. 2007;9(6):223.
[9] WU F, GAO J, KANG J, et al. B Cells in Rheumatoid Arthritis, Pathogenic Mechanisms and Treatment Prospects. Front Immunol. 2021;12:750753.
[10] FLOUDAS A, NETO N, ORR C, et al. Loss of balance between protective and pro-inflammatory synovial tissue T-cell polyfunctionality predates clinical onset of rheumatoid arthritis [published correction appears in Ann Rheum Dis. 2023; 82(2):e53]. Ann Rheum Dis. 2022;81(2): 193-205.
[11] JIANG Q, YANG G, LIU Q, et al. Function and Role of Regulatory T Cells in Rheumatoid Arthritis. Front Immunol. 2021;12:626193.
[12] ZHONG Y, ZHANG W, HONG X, et al. Screening Biomarkers for Systemic Lupus Erythematosus Based on Machine Learning and Exploring Their Expression Correlations With the Ratios of Various Immune Cells. Front Immunol. 2022;13:873787.
[13] YANG Y, CAO Y, HAN X, et al. Revealing EXPH5 as a potential diagnostic gene biomarker of the late stage of COPD based on machine learning analysis. Comput Biol Med. 2023;154:106621.
[14] UNGETHUEM U, HAEUPL T, WITT H, et al. Molecular signatures and new candidates to target the pathogenesis of rheumatoid arthritis. Physiol Genomics. 2010;42A(4):267-282.
[15] WOETZEL D, HUBER R, KUPFER P, et al. Identification of rheumatoid arthritis and osteoarthritis patients by transcriptome-based rule set generation. Arthritis Res Ther. 2014;16(2):R84.
[16] HUBER R, HUMMERT C, GAUSMANN U, et al. Identification of intra-group, inter-individual, and gene-specific variances in mRNA expression profiles in the rheumatoid arthritis synovial membrane. Arthritis Res Ther. 2008;10(4):R98.
[17] BROEREN MG, DE VRIES M, BENNINK MB, et al. Disease-Regulated Gene Therapy with Anti-Inflammatory Interleukin-10 Under the Control of the CXCL10 Promoter for the Treatment of Rheumatoid Arthritis. Hum Gene Ther. 2016;27(3):244-254.
[18] GUO Y, WALSH AM, FEARON U, et al. CD40L-Dependent Pathway Is Active at Various Stages of Rheumatoid Arthritis Disease Progression. J Immunol. 2017;198(11): 4490-4501.
[19] WANG H, YANG F, LUO Z. An experimental study of the intrinsic stability of random forest variable importance measures. BMC Bioinformatics. 2016;17:60.
[20] LIN X, LI C, ZHANG Y, et al. Selecting Feature Subsets Based on SVM-RFE and the Overlapping Ratio with Applications in Bioinformatics. Molecules. 2017;23(1):52.
[21] YANG C, DELCHER C, SHENKMAN E, et al. Machine learning approaches for predicting high cost high need patient expenditures in health care. Biomed Eng Online. 2018; 17(Suppl 1):131.
[22] SANZ H, VALIM C, VEGAS E, et al. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018; 19(1):432.
[23] ZENG D, YE Z, SHEN R, et al. IOBR: Multi-Omics Immuno-Oncology Biological Research to Decode Tumor Microenvironment and Signatures. Front Immunol. 2021;12:687975.
[24] RIVELLESE F, PITZALIS C. Cellular and molecular diversity in Rheumatoid Arthritis. Semin Immunol. 2021;58:101519.
[25] JANG S, KWON EJ, LEE JJ. Rheumatoid Arthritis: Pathogenic Roles of Diverse Immune Cells. Int J Mol Sci. 2022;23(2): 905.
[26] AN Q, RAHMAN S, ZHOU J, et al. A Comprehensive Review on Machine Learning in Healthcare Industry: Classification, Restrictions, Opportunities and Challenges. Sensors (Basel). 2023; 23(9):4178.
[27] TANG H, LIU W, XU Z, et al. Integrated microenvironment-associated genomic profiles identify LRRC15 mediating recurrent glioblastoma-associated macrophages infiltration. J Cell Mol Med. 2021;25(12):5534-5546.
[28] RAY U, PATHOULAS CL, THIRUSANGU P, et al. Exploiting LRRC15 as a Novel Therapeutic Target in Cancer. Cancer Res. 2022;82(9):1675-1681.
[29] PURCELL JW, TANLIMCO SG, HICKSON J, et al. LRRC15 Is a Novel Mesenchymal Protein and Stromal Target for Antibody-Drug Conjugates. Cancer Res. 2018;78(14):4059-4072.
[30] CHEN YJ, CHANG WA, HSU YL, et al. Deduction of Novel Genes Potentially Involved in Osteoblasts of Rheumatoid Arthritis Using Next-Generation Sequencing and Bioinformatic Approaches. Int J Mol Sci. 2017;18(11):2396.
[31] CHEN YJ, CHANG WA, WU LY, et al. Systematic Analysis of Transcriptomic Profile of Chondrocytes in Osteoarthritic Knee Using Next-Generation Sequencing and Bioinformatics. J Clin Med. 2018;7(12):535.
[32] WANG Y, LIU Y, ZHANG M, et al. LRRC15 promotes osteogenic differentiation of mesenchymal stem cells by modulating p65 cytoplasmic/nuclear translocation. Stem Cell Res Ther. 2018;9(1):65.
[33] SINGH P, WANG M, MUKHERJEE P, et al. Transcriptomic and epigenomic analyses uncovered Lrrc15 as a contributing factor to cartilage damage in osteoarthritis. Sci Rep. 2021;11(1):21107.
[34] FERRARI DE ANDRADE L, TAY RE, PAN D, et al. Antibody-mediated inhibition of MICA and MICB shedding promotes NK cell-driven tumor immunity. Science. 2018;359(6383):1537-1542.
[35] STEPHENS HA. MICA and MICB genes: can the enigma of their polymorphism be resolved? Trends Immunol. 2001;22(7): 378-385.
[36] FERRARI DE ANDRADE L, KUMAR S, LUOMA AM, et al. Inhibition of MICA and MICB Shedding Elicits NK-Cell-Mediated Immunity against Tumors Resistant to Cytotoxic T Cells. Cancer Immunol Res. 2020;8(6):769-780.
[37] COURAU T, BONNEREAU J, CHICOTEAU J, et al. Cocultures of human colorectal tumor spheroids with immune cells reveal the therapeutic potential of MICA/B and NKG2A targeting for cancer treatment. J Immunother Cancer. 2019;7(1):74.
[38] ALVES DA SILVA PH, XING S, KOTINI AG, et al. MICA/B antibody induces macrophage-mediated immunity against acute myeloid leukemia. Blood. 2022;139(2):205-216.
[39] LÜ M, XIA B, LI J, et al. MICB microsatellite polymorphism is associated with ulcerative colitis in Chinese population. Clin Immunol. 2006;120(2):199-204.
[40] OTA M, KATSUYAMA Y, KIMURA A, et al. A second susceptibility gene for developing rheumatoid arthritis in the human MHC is localized within a 70-kb interval telomeric of the TNF genes in the HLA class III region. Genomics. 2001;71(3):263-270.
[41] LÓPEZ-ARBESU R, BALLINA-GARCÍA FJ, ALPERI-LÓPEZ M, et al. MHC class I chain-related gene B (MICB) is associated with rheumatoid arthritis susceptibility. Rheumatology (Oxford). 2007;46(3):426-430.