泌尿系结石患者尿有机酸代谢谱特征及关键代谢标志物的筛选

Characteristics of urinary organic acid metabolic profile and screening of key meta- bolic markers in patients with urolithiasis

摘要

目的：系统分析泌尿系结石患者与非结石患者尿有机酸代谢谱的差异，筛选疾病相关特征性代谢标志物及关键信号通路，从代谢层面阐明疾病发生发展的潜在机制，为泌尿系结石的基础研究与干预靶点筛选提供理论依据。方法：选择2024年12月至2025年9月于北京大学人民医院就诊的泌尿系结石患者 99 例与 57 例非泌尿系结石患者（泌尿外科以外未患有泌尿系结石的就诊者），收集其 24 h 尿，通过气相色谱-质谱（gas chromatography-mass spectrometry，GC-MS）技术行非靶向尿有机酸代谢组学检测。采用主成分分析（principal component analysis， PCA）和正交偏最小二乘判别分析（orthogonal partial least squares discriminant analysis，OPLS-DA）组间代谢谱差异及临床协变量的影响；通过火山图、变量重要性投影（variable importance in the projection，VIP）分析筛选差异有机酸；利用 KEGG（Kyoto Encyclopedia of Genes and Genomes）数据库开展代谢通路富集分析；采用机器学习筛选最具代表性的核心代谢物组合，验证有机酸对疾病状态的区分效能。结果：结石组与非结石组患者的尿有机酸代谢呈显著分离趋势（OPLS-DA：$\mathit{R}2\mathit{X}=0.139\mathrm{、}\mathit{R}2\mathit{Y}=0.48\mathrm{、}\mathit{Q}2=0.457$），性别是影响代谢谱的重要协变量，糖尿病、高血压及结石复发状态对尿有机酸谱无显著影响。共筛选出44种差异有机酸（P<0.05,6 种上调、38种下调），VIP 分析确定10 种核心差异代谢物（P<0.001 ，均下调）；差异有机酸主要富集于乙醛酸和二羧酸代谢等能量代谢通路及色氨酸代谢等氨基酸代谢通路。机器学习模型中梯度提升机（gradient boosting machine，GBM）、随机森林（random forest， RF）效能最优，平均曲线下面积（area under the curve，AUC）为0.996，筛选出丙酰甘氨酸、5-羟基吲哚-3-乙酸、犬尿喹啉酸、乳清酸、富马酸 5 种核心代谢物为最佳组合，基于该组合构建的列线图模型 AUC=0.997 ，校准曲线校正偏倚后与理想曲线高度吻合，平均绝对误差为 0.022 ，能够高效区分结石与非结石人群，提示这些代谢物可以作为泌尿系结石机制研究的潜在靶点。结论：泌尿系结石患者存在显著尿有机酸代谢紊乱，代谢异常主要涉及能量及氨基酸代谢通路，其紊乱与疾病发病机制密切相关。丙酰甘氨酸、5-羟基吲哚-3-乙酸、犬尿喹啉酸、乳清酸、富马酸5种核心代谢物可以有效地区分结石与非结石人群，其可以作为代表性的代谢标志物为后续机制探索与靶向干预提供新靶点。

Abstract

Objective：To systematically compare the differences in urinary organic acid metabolic pro-files between patients with urinary calculi and non-calculi individuals，to screen disease-specific charac- teristic metabolic biomarkers and key signaling pathways，and to elucidate the potential mechanism underlying the occurrence and progression of urinary calculi at the metabolic level，so as to provide a theoretical basis for basic research and screening of intervention targets for urinary calculi. Methods：A total of 99 patients with urinary calculi and 57 non-calculi subjects were enrolled.Twenty-four-hour urine samples were collected and subjected to untargeted urinary organic acid metabolomics detection by gas chromatography-mass spectrometry（GC-MS）.Principal component analysis（PCA）and orthogonal partial least squares discriminant analysis（OPLS-DA）were used to analyze the differences in metabolic profiles between the groups and the effects of clinical covariates.Volcano plot and variable importance in the pro- jection（VIP）analysis were applied to screen differential organic acids.The Kyoto Encyclopedia of Genes and Genomes（KEGG）database were used for metabolic pathway enrichment analysis.Machine learning algorithms were adopted to screen the most representative core metabolite combinations and verify the discriminatory efficacy of organic acids for disease status. Results：The urinary organic acid metabolic profiles showed a significant separation trend between the calculi group and the non-calculi group（OPLS-DA：$\mathit{R}2\mathit{X}=0.139,\mathit{R}2\mathit{Y}=0.48,\mathit{Q}2=0.457$）.Gender was an important covariate affecting metabolic profiles，while diabetes，hypertension and stone recurrence status exerted no significant influence.A total of 44 differential organic acids were screened（P<0.05），including 6 upregulated and 38 downregulated metabolites.Ten core differential metabolites were identified via VIP analysis（P<0.001 ，all downregu- lated）.Differential organic acids were mainly enriched in energy metabolism pathways such as glyoxylate and dicarboxylate metabolism，as well as amino acid metabolism pathways including tryptophan metabo- lism.Among machine learning models，Gradient Boosting Machine and Random Forest exhibited the opti- mal efficacy with an average area under the curve（AUC）of 0.996.Five core metabolites including pro- pionylglycine，5-hydroxyindole-3-acetic acid，kynurenic acid，orotic acid and fumaric acid were identi- fied as the optimal combination.The nomogram model constructed based on this combination yielded an AUC of 0.997 ；after calibration bias correction，the calibration curve was highly consistent with the ideal curve，with a mean absolute error of 0.022 ，which could effectively distinguish calculi patients from non- calculi individuals. Conclusion：Patients with urinary calculi present obvious disorders in urinary organic acid metabolism，which are mainly involved in energy and amino acid metabolic pathways and closely associated with the pathogenesis of urinary calculi.The five core metabolites（propionylglycine，5-hydroxyindole-3-acetic acid，kynurenic acid，orotic acid and fumaric acid）can effectively distinguish calculi from non-calculi populations，and can serve as representative metabolic biomarkers to provide no- vel targets for subsequent mechanism exploration and targeted intervention.

关键词

泌尿系结石 / 尿有机酸 / 代谢组学 / 泌尿外科

Key words

Urolithiasis / Urinary organic acids / Metabolomics / Urology

参考文献

[1] Shoag J, Tasian GE, Goldfarb DS, et al. The new epidemiology of nephrolithiasis[J]. Adv Chronic Kidney Dis, 2015, 22 (4): 273-278.

[2] Ziemba JB, Matlaga BR. Epidemiology and economics of nephroli- thiasis[J]. Investig Clin Urol, 2017, 58(5): 299.

[3] Sorokin I, Mamoulakis C, Miyazawa K, et al. Epidemiology of stone disease across the world[J]. World J Urol, 2017, 35(9): 1301-1320.

[4] Tang X, Lieske JC. Acute and chronic kidney injury in nephroli- thiasis[J]. Curr Opin Nephrol Hypertens, 2014, 23(4): 385-390.

[5] Bargagli M, Scoglio M, Howles SA, et al. Kidney stone disease: Risk factors,pathophysiology and management[J]. Nat Rev Nephrol, 2025, 21(11):794-808.

[6] Kim HN, Kim JH, Chang Y, et al. Gut microbiota and the preva- lence and incidence of renal stones[J]. Sci Rep, 2022, 12(1): 3732.

[7] Lai Y, Zheng H, Sun X, et al. The advances of calcium oxalate calculi associated drugs and targets[J]. Eur J Pharmacol, 2022, 935: 175324.

[8] Peerapen P, Thongboonkerd V. Kidney stone prevention[J]. Adv Nutr, 2023, 14(3):555-569.

[9] Zhang XZ, Lei XX, Jiang YL, et al. Application of metabolomics in urolithiasis:The discovery and usage of succinate[J]. Sig Transduct Target Ther, 2023, 8(1): 41.

[10] Duan X, Zhang T, Ou L, et al. ¹ H NMR-based metabolomic study of metabolic profiling for the urine of kidney stone patients[J]. Urolithiasis, 2020, 48(1):27-35.

[11] Kaneko K, Kobayashi R, Yasuda M, et al. Comparison of matrix proteins in different types of urinary stone by proteomic analysis using liquid chromatography-tandem mass spectrometry[J]. Int J Urol, 2012, 19(8):765-772.

[12] Boonla C, Tosukhowong P, Spittau B, et al. Inflammatory and fi- brotic proteins proteomically identified as key protein constituents in urine and stone matrix of patients with kidney calculi[J]. Clin Chim Acta, 2014, 429:81-89.

[13] Wang Z, Zhang Y, Zhang J, et al. Recent advances on the mechanisms of kidney stone formation(Review)[J]. Int J Mol Med, 2021, 48(2): 149.

[14] Bolanos-Palmieri P, Schenk H, Bähre H, et al. Kynurenine path- way dysregulation impairs podocyte morphology and bioenergetics in vitro and leads to glomerular dysfunction[J]. FASEB J, 2025, 39(22):e71228.

[15] Liu JR, Miao H, Deng DQ, et al. Gut microbiota-derived trypto- phan metabolism mediates renal fibrosis by aryl hydrocarbon recep- tor signaling activation[J]. Cell Mol Life Sci, 2021, 78 (3): 909-922.

[16] Agudelo LZ, Ferreira DMS, Cervenka I, et al. Kynurenic acid and Gpr35 regulate adipose tissue energy homeostasis and inflam- mation[J]. Cell Metab, 2018, 27(2):378-392.e5.

[17] Zakrocka I, Załuska W. Kynurenine pathway in kidney diseases[J]. Pharmacol Rep, 2022, 74(1):27-39.

[18] Sadaf H, Raza SI, Hassan SW. Role of gut microbiota against cal- cium oxalate[J]. Microb Pathog, 2017, 109:287-291.

[19] Song Q, Song C, Chen X, et al. Oxalate regulates crystal-cell ad- hesion and macrophage metabolism via JPT2/PI3K/AKT signaling to promote the progression of kidney stones[J]. J Pharm Anal, 2024, 14(6): 100956.

[20] Tavasoli S, Alebouyeh M, Naji M, et al. Association of intestinal oxalate-degrading bacteria with recurrent calcium kidney stone for- mation and hyperoxaluria:A case-control study[J]. BJU Int, 2020, 125(1):133-143.

[21] Cao C, Li F, Ding Q, et al. Potassium sodium hydrogen citrate intervention on gut microbiota and clinical features in uric acid stone patients[J]. Appl Microbiol Biotechnol, 2024, 108 (1): 51.

[22] Gao H, Lin J, Xiong F, et al. Urinary microbial and metabolomic profiles in kidney stone disease[J]. Front Cell Infect Microbiol, 2022, 12: 953392.