整合生物信息学和机器学习算法筛选心房颤动炎性标志物与验证^*

Integrating Bioinformatics and Machine Learning Algorithms to Screen Inflammatory Biomarkers for Atrial Fibrillation and Experimental Validation

摘要

目的 通过整合生物信息学和机器学习算法及实验研究，筛选并鉴定心房颤动（AF）的潜在生物标志物。方法 从GEO和MSigDB数据库中获取AF转录组学并筛选差异表达基因（DEGs），从炎症相关基因集（IRGs）。取交集形成DE-IRGs基因集。采用机器学习算法过滤AF相关DE-IRGs基因后获得特征基因。取12只SPF级雄性SD大鼠，分为对照组和AF组。其中AF组大鼠连续5周尾静脉注射乙酰胆碱（66 μg/mL）和氯化钙（10 mg/kg）构建AF模型；HE染色观察心房肌细胞形态特征、RT-qPCR和免疫组织化学染色测定特征基因表达mRNA和蛋白质变化。结果 训练集发现119个DEGs，IRGs与两个共表达模块相关性最高（r=0.6和0.56，P<0.000 1）。DEGs与IRGs取交集获得9个DE-IRGs；SVM-RFE和RF算法筛选出的4个特征基因5羟色胺受体2B基因（5-hydroxytryptamine receptor 2B gene,HTR2B）、潜在转化生长因子β结合蛋白2基因（latent-transforming growth factor beta-binding protein 2 gene,LTBP2）、基质重塑关联蛋白5基因（matrix remodeling associated protein 5 gene,MXRA5）和转化生长因子β诱导蛋白基因（transforming growth factor β-induced protein gene, TGFBI）在训练集和验证集AF组均呈高表达且诊断效能较高（AUC>0.77）。AF大鼠心电图肢体导联出现大量AF特征性f波。HE染色显示AF组心房肌细胞排列紊乱、炎性细胞浸润。TUNEL荧光检测AF组凋亡率为（55.34±4.29）%，而对照组为（8.69±3.12）%（P=0.000 1）。LTBP2和TGFBI基因相对表达量在AF组（4.97±4.20，2.62±1.85）和对照组（1.12±0.21，1.18±0.77）间差异存在统计学意义（P=0.013 7，P=0.044 4）。免疫组织化学染色检测AF组LTBP2和TGFBI蛋白阳性率分别为（36.50±1.31）%和（27.39±4.57）%，而对照组分别是（22.95±2.62）%和（18.26±3.70）%（P=0.000 8，P=0.048 5）。结论 特征基因LTBP2和TGFBI在AF中高表达，且较有诊断价值的生物标志物。

Abstract

Objective To identify potential biomarkers of atrial fibrillation (AF) using bioinformatics, machine learning (ML) and experimental methods. Methods AF transcriptomic data were obtained from the GEO and MSigDB databases to screen for differentially expressed genes (DEGs) and inflammation-related gene sets (IRGs). The overlap between DEGs and IRGs was used to identify the DE-IRGs set. Two machine learning algorithms were used to filter AF-related DE-IRGs. Twelve male SD rats were randomly divided into a control group and an AF group. Rats received acetylcholine (66 μg/mL) and calcium chloride (10 mg/kg) via tail vein injection for five consecutive weeks to establish the AF model. Morphological characteristics of atrial myocytes were assessed with HE staining, while RT-qPCR and immunohistochemistry (IHC) were used to measure changes in mRNA and protein levels. Results In the training set, 119 DEGs were identified, with IRGs showing the highest correlation with two co-expression modules (r = 0.60 and 0.56, P < 0.000 1). The intersection of DEGs and IRGs yielded nine DE-IRGs. The SVM-RFE and RF algorithms identified 5-hydroxytryptamine receptor 2B gene ( HTR2B), latent-transforming growth factor beta-binding protein 2 gene (LTBP2), matrix remodeling associated protein 5 gene (MXRA5), and transforming growth factor β-induced protein gene (TGFBI) as highly expressed in both the training and test sets in AF groups with high diagnostic efficacy (AUC > 0.77). The electrocardiogram limb leads of AF rats showed numerous f-waves. HE staining revealed disorganized atrial myocyte arrangement and inflammatory cell infiltration in the AF group. The TUNEL fluorescence assay showed an apoptosis rate of (55.34 ± 4.29)% in the AF group, compared to (8.69 ± 3.12)% in the control group ( P = 0.000 1). There were statistically significant differences in the relative expression levels of LTBP2 and TGFBI between the AF group (4.97 ± 4.20, 2.62 ± 1.85) and the control group (1.12 ± 0.21, 1.18 ± 0.77) (P = 0.013 7, P = 0.044 4). IHC revealed positive rates of LTBP2 and TGFBI proteins in the AF group of (36.50 ± 1.31)% and (27.39 ± 4.57)%, respectively, compared to (22.95 ± 2.62)% and (18.26 ± 3.70)% in the control group (P = 0.000 8, P = 0.048 5). Conclusion The characteristic genes LTBP2 and TGFBI are highly expressed in AF and serve as valuable diagnostic biomarkers.

关键词

心房颤动 / 机器学习算法 / 生物标志物 / 炎症反应

Key words

Atrial fibrillation / Machine learning algorithms / Biomarkers / Inflammatory response

参考文献

[1] BRUNDEL B, AI X, HILLS M T, et al. Atrial fibrillation. Nat Rev Dis Primers, 2022, 8(1): 21. doi: 10.1038/s41572-022-00347-9.

[2] BORIANI G, LAROCHE C, DIEMBERGER I, et al. Asymptomatic atrial fibrillation: clinical correlates, management, and outcomes in the EORP-AF Pilot General Registry. Am J Med, 2015, 128(5): 509-518.e2. doi: 10.1016/j.amjmed.2014.11.026.

[3] WITKOWSKI M, BISSINGER A, GRYCEWICZ T, et al. Asymptomatic atrial fibrillation in patients with atrial fibrillation and implanted pacemaker. Int J Cardiol, 2017, 227: 583-588. doi: 10.1016/j.ijcard.2016.10.097.

[4] WEGNER F K, KOCHHÄUSER S, ELLERMANN C, et al. Prospective blinded evaluation of the smartphone-based AliveCor Kardia ECG monitor for atrial fibrillation detection: The PEAK-AF study. Eur J Intern Med, 2020, 73: 72-75. doi: 10.1016/j.ejim.2019.11.018.

[5] Van Den BERG M P, MULDER B A, KLAASSEN S H C, et al. Heart failure with preserved ejection fraction, atrial fibrillation, and the role of senile amyloidosis. Eur Heart J, 2019, 40(16): 1287-1293. doi: 10.1093/eurheartj/ehz057.

[6] FUHER A N, BORNE R, CUNNINGHAM J. Diagnosing atypical flutter in the post-atrial fibrillation ablation patient: a case report. Clin Pract Cases Emerg Med, 2023, 7(2): 106-109. doi: 10.5811/cpcem.1413.

[7] WU L D, LI F, CHEN J Y, et al. Analysis of potential genetic biomarkers using machine learning methods and immune infiltration regulatory mechanisms underlying atrial fibrillation. BMC Med Genomics, 2022, 15(1): 64. doi: 10.1186/s12920-022-01212-0.

[8] DEMERS J, TON A T, HUYNH F, et al. Atrial electrical remodeling in mice with cardiac-specific overexpression of angiotensin Ⅱ type 1 receptor. J Am Heart Assoc, 2022, 11(8): e023974. doi: 10.1161/jaha.121.023974.

[9] ZHENG L, JIA X, ZHANG C, et al. Angiotensin Ⅱ in atrial structural remodeling: the role of Ang Ⅱ/JAK/STAT3 signaling pathway. Am J Transl Res, 2015, 7(6): 1021-1031.

[10] MURAKATA Y, YAMAGAMI F, MURAKOSHI N, et al. Electrical, structural, and autonomic atrial remodeling underlies atrial fibrillation in inflammatory atrial cardiomyopathy. Front Cardiovasc Med, 2022, 9: 1075358. doi: 10.3389/fcvm.2022.1075358.

[11] SMORODINOVA N, LANTOVÁ L, BLÁHA M, et al. Bioptic study of left and right atrial interstitium in cardiac patients with and without atrial fibrillation: interatrial but not rhythm-based differences. PLoS One, 2015, 10(6): e0129124. doi: 10.1371/journal.pone.0129124.

[12] ALTIERI D I, ETZION Y, ANDERSON H D. Cannabinoid receptor agonist attenuates angiotensin Ⅱ-induced enlargement and mitochondrial dysfunction in rat atrial cardiomyocytes. Front Pharmacol, 2023, 14: 1142583. doi: 10.3389/fphar.2023.1142583.

[13] LANGFELDER P, HORVATH S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics, 2008, 9(1): 559. doi: 10.1186/1471-2105-9-559.

[14] HÄNZELMANN S, CASTELO R, GUINNEY J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics, 2013, 14: 7. doi: 10.1186/1471-2105-14-7.

[15] WU T, HU E, XU S, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation (Camb), 2021, 2(3): 100141. doi: 10.1016/j.xinn.2021.100141.

[16] LIU P, SUN H, ZHOU X, et al. CXCL12/CXCR4 axis as a key mediator in atrial fibrillation via bioinformatics analysis and functional identification. Cell Death Dis, 2021, 12(9): 813. doi: 10.1038/s41419-021-04109-5.

[17] SZILÁGYI J, SÁGHY L. Atrial remodeling in atrial fibrillation. Comorbidities and markers of disease progression predict catheter ablation outcome. Curr Cardiol Rev, 2021, 17(2): 217-229. doi: 10.2174/1573403x16666200721153620.

[18] LI C Y, ZHANG J R, HU W N, et al. Atrial fibrosis underlying atrial fibrillation (review). Int J Mol Med, 2021, 47(3): 9. doi: 10.3892/ijmm.2020.4842.

[19] MOLINA C E, ABU-TAHA I H, WANG Q, et al. Profibrotic, electrical, and calcium-handling remodeling of the atria in heart failure patients with and without atrial fibrillation. Front Physiol, 2018, 9: 1383. doi: 10.3389/fphys.2018.01383.

[20] WANG W W, ZHANG F L, CHEN J H, et al. Telmisartan reduces atrial arrhythmia susceptibility through the regulation of RAS-ERK and PI3K-Akt-eNOS pathways in spontaneously hypertensive rats. Can J Physiol Pharmacol, 2015, 93(8): 657-665. doi: 10.1139/cjpp-2014-0416.

[21] LIU Z, FINET J E, WOLFRAM J A, et al. Calcium/calmodulin-dependent protein kinase Ⅱ causes atrial structural remodeling associated with atrial fibrillation and heart failure. Heart Rhythm, 2019, 16(7): 1080-1088. doi: 10.1016/j.hrthm.2019.01.013.

[22] YU J F, DONG Q, DU Y M. Interleukin-6: molecular mechanisms and therapeutic perspectives in atrial fibrillation. Curr Med Sci, 2025, 45(2): 157-168. doi: 10.1007/s11596-025-00021-7.

[23] LARSEN B C, Al-BADRI K H A, MELLEMKJÆR S, et al. Inflammatory-mediated atrial cardiomyopathy diagnosed using multimodality imaging and successfully treated with prednisolone: a case report. Eur Heart J Case Rep, 2024, 8(4): ytae167. doi: 10.1093/ehjcr/ytae167.

[24] LAKIN R, POLIDOVITCH N, YANG S, et al. Cardiomyocyte and endothelial cells play distinct roles in the tumour necrosis factor (TNF)-dependent atrial responses and increased atrial fibrillation vulnerability induced by endurance exercise training in mice. Cardiovasc Res, 2023, 119(16): 2607-2622. doi: 10.1093/cvr/cvad144.

[25] 李春兰, 裴丽娟, 侯鹏. 心房颤动相关的生物学标记物. 中国心血管杂志, 2020, 25(4): 397-400. doi: 10.3969/j.issn.1007-5410.2020.04.020.

[26] LI C L, PEI L J, HOU P. Biomarkers related to atrial fibrillation. Chiese Journal of Cardiovascular Medicine, 2020, 25(4): 397-400. doi: 10.3969/j.issn.1007-5410.2020.04.020.

[27] 吴希泽, 李越, 潘嘉祥, 等. 基于生物信息学探讨中国人群发生心房颤动的机制及潜在生物标志物. 浙江大学学报(医学版), 2024, 53(5): 593-603. doi: 10.3724/zdxbyxb-2024-0027.

[28] WU X, LI Y, PAN J, et al. Pathogenesis and potential diagnostic biomarkers of atrial fibrillation in Chinese population: a study based on bioinformatics. Journal of Zhejiang University (Medical Sciences), 2024, 53(5): 593-603. doi: 10.3724/zdxbyxb-2024-0027.

[29] ROBERTSON I B, HORIGUCHI M, ZILBERBERG L, et al. Latent TGF-β-binding proteins. Matrix Biol, 2015, 47: 44-53. doi: 10.1016/j.matbio.2015.05.005.

[30] PANG X F, LIN X, DU J J, et al. LTBP2 knockdown by siRNA reverses myocardial oxidative stress injury, fibrosis and remodelling during dilated cardiomyopathy. Acta Physiol (Oxf), 2020, 228(3): e13377. doi: 10.1111/apha.13377.

[31] WANG Y, HUA D, ZHAO K, et al. LTBP2 knockdown ameliorates cardiac fibrosis and apoptosis via attenuating NF-κB signaling pathway and oxidative stress in mice with cardiac hypertrophy. Advanced Therapeutics, 2024, 7(8): 2300451. doi: 10.1002/adtp.202300451.

[32] CHEN H X, YANG Z Y, HOU H T, et al. Novel mutations of TCTN3/LTBP2 with cellular function changes in congenital heart disease associated with polydactyly. J Cell Mol Med, 2020, 24(23): 13751-13762. doi: 10.1111/jcmm.15950.

[33] HORIGUCHI M, TODOROVIC V, HADJIOLOVA K, et al. Abrogation of both short and long forms of latent transforming growth factor-β binding protein-1 causes defective cardiovascular development and is perinatally lethal. Matrix Biol, 2015, 43: 61-70. doi: 10.1016/j.matbio.2015.03.006.

[34] ZHOU M, XU Y, ZHANG L, et al. Comprehensive bioinformatics analysis of hub genes in ischemic heart failure and atrial fibrillation. Front Cardiovasc Med, 2025, 12: 1499065. doi: 10.3389/fcvm.2025.1499065.

[35] ZHAO L, XU Y, TAO L, et al. Oxymatrine inhibits transforming growth factor β1 (TGF-β1)-induced cardiac fibroblast-to-myofibroblast transformation (FMT) by mediating the notch signaling pathway in vitro . Med Sci Monit, 2018, 24: 6280-6288. doi: 10.12659/msm.910142.

[36] LIU C, ZHOU D, ZHANG Q, et al. Transcription factor EB (TFEB) improves ventricular remodeling after myocardial infarction by inhibiting Wnt/β-catenin signaling pathway. PeerJ, 2023, 11: e15841. doi: 10.7717/peerj.15841.

[37] REN Z, LI T, LIU X, et al. Transforming growth factor-beta 1 enhances discharge activity of cortical neurons. Neural Regen Res, 2025, 20(2): 548-556. doi: 10.4103/nrr.Nrr-d-23-00756.

[38] HAN D H, SHIN M K, SUNG J S, et al. miR-335-3p attenuates transforming growth factor beta 1-induced fibrosis by suppressing thrombospondin 1. PLoS One, 2024, 19(10): e0311594. doi: 10.1371/journal.pone.0311594.

[39] POULSEN E T, RUNAGER K, NIELSEN N S, et al. Proteomic profiling of TGFBI-null mouse corneas reveals only minor changes in matrix composition supportive of TGFBI knockdown as therapy against TGFBI-linked corneal dystrophies. FEBS J, 2018, 285(1): 101-114. doi: 10.1111/febs.14321.

[40] SCIRIHA G G, BORG J, SULTANA J, et al. Transcriptome analysis of TGFBI knockdown vs normal corneal epithelial cells: implications for TGFBI corneal dystrophy treatment. Biochem Genet, 2026, 64(3): 3451-3480. doi: 10.1007/s10528-025-11191-3.

[41] CHEN Z, WANG J, PENG P, et al. Hypoxia-induced TGFBI maintains glioma stem cells by stabilizing EphA2. Theranostics, 2024, 14(15): 5778-5792. doi: 10.7150/thno.95141.

基金资助

*昆明市卫生健康委员会卫生科研课题项目(2023-03-01-007)