Analysis of diagnostic biomarkers for ischemic stroke and experimental validation of targeted cuproptosis related genes

doi:10.12307/2025.957

Abstract

Abstract: BACKGROUND: Studies have shown that immune cells are involved in all processes of ischemic stroke, in which cuproptosis also plays a key role.
OBJECTIVE: To screen diagnostic biomarkers related to the progression of ischemic stroke through bioinformatics, and analyze and validate cuproptosis-related genes closely related to the occurrence and development of ischemic stroke.
METHODS: The GSE16561 microarray was obtained from the GEO database, containing data from 39 cases of ischemic stroke (ischemic stroke group) and 24 controls (control group). Differentially expressed genes from the ischemic stroke microarray data were analyzed. Gene ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were performed. By using LASSO and Random Forest methods, key genes affecting the occurrence and development of ischemic stroke were screened, and a diagnostic model was established and validated. Differential gene analysis was performed through immune cell infiltration and weighted gene co expression network. The differentially expressed immune-related genes were intersected with cuproptosis genes to obtain the hub genes related to cuproptosis immunity. In vitro cell experiments were conducted to divide rat hippocampal neurons into a normal group and an ischemic stroke group, and qPCR experiments were performed to verify the results.
RESULTS AND CONCLUSION: (1) 573 differentially expressed genes were obtained by differential analysis. Differentially expressed genes were mainly enriched in biological processes, such as positive regulation of immune response, and signaling pathways such as lipid and atherosclerosis. (2) Machine learning methods were used to screen diagnostic genes such as MFN2, PKM2, CREG1, and FOXO3A, which may have some diagnostic value for ischemic stroke. (3) Immune infiltration analysis revealed resting plasma cells, NK cells, macrophages, etc., indicating that immune cells play a certain role in the pathogenesis of ischemic stroke. (4) Weighted gene co-expression network analysis combined with immune infiltration analysis obtained 118 key module genes, which were intersected with cuproptosis genes to obtain 2 cuproptosis and immune characteristic genes. The correlation analysis between four diagnostic genes and Hub genes showed that the expression of FOXO3A and MFN2, PKM2 and BCL2L1, MTF1 and MFN2, ATP7B and BCL2L1 were correlated. (5) The qPCR results showed significant differences in the genes MTF1 and ATP7B between the ischemic stroke group and the blank group. To conclude, ATP7B and MTF1 can serve as characteristic genes for cuproptosis in ischemic stroke. It is possible to improve ischemic stroke by intervening in ATP7B and MTF1 to regulate cuproptosis.

Key words: ischemic stroke, machine learning, bioinformatics, immune infiltration, cuproptosis, cellular experiments, ATP7B, MTF1

CLC Number:

Chen Ying, Guo Xiaojing, Mo Xueni, Ma Wei, Wu Shangzhi, Li Xiangling, Xie Tingting. Analysis of diagnostic biomarkers for ischemic stroke and experimental validation of targeted cuproptosis related genes[J]. Chinese Journal of Tissue Engineering Research, 2025, 29(35): 7562-7570.

Figures/Tables 8

2.1 芯片数据规范化处理后差异表达基因的识别首先对GSE16561数据集进行归一化处理，然后通过“limma”包对规范化后的芯片进行差异分析，获得573个差异表达基因，其中141个差异表达基因表达下调，432个差异表达基因表达上调，进一步绘制的火山图和热图，见图1。 2.2 差异基因富集分析 GO富集分析揭示，差异表达基因显著集中于免疫反应的正调控、分泌颗粒膜以及免疫受体活性的调节等生物学途径。而KEGG通路分析则显示，这些差异表达基因主要关联于脂质代谢与动脉粥样硬化、T细胞受体信号传导、催产素信号通路以及cGMP-PKG信号传导途径等相关信号网络，见图2。 2.3 机器学习模型的构建与验证为了构建并验证机器学习模型以识别影响缺血性脑卒中的关键诊断性基因，采用了2种机器学习算法。LASSO回归分析从具有统计意义的变量中精选出15个预测基因(见图3A)，而Random Forest算法则筛选出10个预测基因(见图3B)。图4可见，通过绘制VENN图，确定了4个共同基因：MFN2、PKM2、CREG1和FOXO3A(见图4B)。利用“rms”包构建了包含这4个基因的缺血性脑卒中诊断列线图(见图4A)。校准曲线显示，模型预测的风险与真实患病风险之间差异甚微，表明该列线图模型在预测缺血性脑卒中发病方面具有较高的准确性(见图4C)。DCA分析揭示，当风险阈值在0-1范围内时，这4个基因的组合能为患者带来显著的临床效益(见图5A)。验证集的分析结果与训练集高度一致(见图5B)，这强化了这4个基因与缺血性脑卒中发病之间的紧密联系。ROC分析结果显示，MFN2 (AUC：0.757)、PKM2(AUC：0.718)、CREG1(AUC：0.838)和FOXO3A(AUC：0.778)的AUC值均超过0.7(见图5C)，验证集的数据分析结果与训练集相似(见图5D)，这进一步证实了这4个基因在缺血性脑卒中诊断中的重要价值。 2.4 免疫细胞浸润分析利用“CIBERSORT”软件包进行免疫细胞浸润的深入分析后，发现疾病组与对照组之间存在13种免疫细胞表达的显著差异(P值均< 0.05)。值得注意的是，在疾病组中，浆细胞、NK细胞静息以及巨噬细胞M0等的表达水平相对较低，见图6。 2.5 加权基因共同表达网络的构建基于GEO数据库中GSE16561数据集的聚类分析，构建了加权基因共表达网络。在此过程中，首先识别并排除了离群样本，见图7A。为了确定合适的软阈值β，设定了R2 > 0.9且平均连接性较高的标准，最终选定β值为16，相关结果展示在图7B中。采用动态树切割法成功鉴定出7个核心模块，见图7C。为了深入探究这些模块与免疫细胞浸润之间的关系进行了相关性分析，结果显示灰色模块与嗜酸性粒细胞之间存在高度正相关(r=0.58，P=5×10??)，而棕色模块则与巨噬细胞M0呈显著正相关(r=0.58，P=7×10-?)。基于这些发现，进一步筛选出118个关键模块基因，见图8。 2.6 铜死亡相关基因的筛选及相关性分析铜死亡相关基因与选定的118个关键模块基因进行交叉比对，鉴别出了2个核心基因MTF1和ATP7B，它们既与铜死亡机制相关，又与免疫活动紧密相连，因此被标记为Hub基因。这一发现通过VENN图(图9A)得以直观呈现。进一步的研究表明，MTF1和ATP7B在基因表达水平上展现出明显的差异性，见图9B，C。Hub基因与4个诊断性基因的相关性分析结果显示FOXO3A与MFN2(r=0.75)、PKM2与BCL2L1(r=0.74)、MTF1与MFN2(r=0.62)、ATP7B与BCL2L1(r=0.57)的表达具有相关性，见图10。 2.7 体外细胞实验验证结果 qPCR 结果显示，与空白组相比，缺血性脑卒中组基因MTF1表达下调，ATP7B表达上调；MTF1、ATP7B在缺血性脑卒中组和空白组间表达存在显著差异 (P < 0.05)，见图11。 "

References

[1] TERRILL AL. Mental Health Issues Poststroke: Underrecognized and Undertreated. Stroke. 2023;54(6):1528-1530.
[2] 唐春花,郭露,李琼,等. 2022年全球卒中数据报告解读[J].诊断学理论与实践,2023,22(3):238-246.
[3] KAHLSON MA, DIXON SJ. Copper-induced cell death. Science. 2022; 375(6586):1231-1232.
[4] REN X, PAN C, PAN Z, et al. Knowledge mapping of copper-induced cell death: A bibliometric study from 2012 to 2022. Medicine (Baltimore). 2022;101(45):e31133.
[5] XUE Q, KANG R, KLIONSKY DJ, et al. Copper metabolism in cell death and autophagy. Autophagy. 2023;19(8):2175-2195.
[6] CHANG W, LI P. Copper and Diabetes: Current Research and Prospect.Mol Nutr Food Res. 2023;67(23):e2300468.
[7] YANG S, LI Y, ZHOU L, et al. Copper homeostasis and cuproptosis in atherosclerosis: metabolism, mechanisms and potential therapeutic strategies. Cell Death Discov. 2024;10(1):25.
[8] CHEN L, MIN J, WANG F. Copper homeostasis and cuproptosis in health and disease. Signal Transduct Target Ther. 2022;7(1):378.
[9] 卓桂锋,陈炜,朱健敏，等. 血管性痴呆铜死亡关键差异基因的生物信息学分析及防治中药筛选[J]. 中草药,2023,54(21):7120-7129.
[10] HE YJ, CONG L, LIANG SL, et al. Discovery and validation of Ferroptosis-related molecular patterns and immune characteristics in Alzheimer’s disease. Front Aging Neurosci. 2022;14:1056312.
[11] FAN J, CHEN M, CAO S, et al. Identification of a ferroptosis-related gene pair biomarker with immune infiltration landscapes in ischemic stroke:a bioinformatics-based comprehensive study. BMC Genomics. 2022;23(1):59.
[12] 王咪,马书杰,刘杨,等.缺血性脑卒中铁死亡特征基因NFE2L2的鉴定与验证[J].中国组织工程研究,2025,29(7):1466-1474.
[13] XING L, WANG Z, HAO Z, et al. Cuproptosis in stroke: focusing on pathogenesis and treatment. Front Mol Neurosci. 2024;17:1349123.
[14] LAI MC, WANG D, LIN ZX, et al. Small Molecule Copper and Its Relative Metabolites in Serum of Cerebral Ischemic Stroke Patients. J Stroke Cerebrovasc Dis. 2016;25(1):214-219.
[15] JIANG Y, WANG LP, DONG XH, et al. Trace Amounts of Copper in Drinking Water Aggravate Cerebral Ischemic Injury via Impairing Endothelial Progenitor Cells in Mice. CNS Neurosci Ther. 2015;21(8):677-680.
[16] SHI L, SUN Z, SU W, et al. Treg cell-derived osteopontin promotes microglia-mediated white matter repair after ischemic stroke. Immunity. 2021;54(7):1527-1542.
[17] ZHANG X, LEI B, YUAN Y, et al. Brain control of humoral immune responses amenable to behavioural modulation. Nature. 2020; 581(7807):204-208.
[18] JIN WN, SHI SX, LI Z, et al. Depletion of microglia exacerbates postischemic inflammation and brain injury. J Cereb Blood Flow Metab. 2017;37(6):2224-2236.
[19] 董一蕾,李珺媛,张业昊,等. 缺血性脑卒中免疫炎症机制及中医药干预研究进展[J].世界中医药,2024,19(2):291-297.
[20] JIANG M, YIN P, BAI X, et al. Proinflammatory and Anti-inflammatory Genes in Stroke Pathogenesis. Curr Pharm Des. 2020;26(34):4220-4233.
[21] SCHALLER B. Prospects for the future: the role of free radicals in the treatment of stroke. Free Radic Biol Med. 2005;38(4):411-425.
[22] XU XH, SHAH PK, FAURE E, et al. Toll-like receptor-4 is expressed by macrophages in murine and human lipid-rich atherosclerotic plaques and upregulated by oxidized LDL. Circulation. 2001;104(25):3103-3108.
[23] MA W, YANG Y, XIE T, et al. Treatment of Ischemic Stroke With Wenyang Huayu Formula: Network Pharmacology Analysis and Experimental Validation. Natural Product Communications. 2023;18(2): 1934578X-2211473X.
[24] YAMAMOTO S, MATSUI A, OHYAGI M, et al. In Vitro Generation of Brain Regulatory T Cells by Co-culturing With Astrocytes. Front Immunol. 2022;13:960036.
[25] SHICHITA T, SUGIYAMA Y, OOBOSHI H, et al. Pivotal role of cerebral interleukin-17-producing gammadeltaT cells in the delayed phase of ischemic brain injury. Nat Med. 2009;15(8):946-950.
[26] RAYASAM A, KIJAK JA, KISSEL L, et al. CXCL13 expressed on inflamed cerebral blood vessels recruit IL-21 producing T(FH) cells to damage neurons following stroke. J Neuroinflammation. 2022;19(1):125.
[27] KHAN S. Artificial Intelligence and Machine Learning in Clinical Medicine. N Engl J Med. 2023;388(25):2398.
[28] CHOI B, ZHENG W. Copper transport to the brain by the blood-brain barrier and blood-CSF barrier. Brain Res. 2009;1248:14-21.
[29] TALUKDER M, BI SS, JIN HT, et al. Cadmium induced cerebral toxicity via modulating MTF1-MTs regulatory axis. Environ Pollut. 2021;285: 117083.