Zhang Pan, Tuo Xiaofeng, Jiang Jiong, Zhang Yue, Zhao Juhui, Deng Chengzhao, Zhao Gang, Cheng Yan, Song Lingqin, Yang Yan, Guo Ruochun, Zhang Huan, Zhao Hongli, Ma Shiyang, Li Lu, Shi Haitao
Department of Gastroenterology, The Second Affiliated Hospital of Xi'an Jiaotong University, No. 157 Xiwu Road, Xi'an, 710004, Shaanxi, China.
Shaanxi Key Laboratory of Gastrointestinal Motility Disorders, Xi'an, 710004, Shaanxi, China.
J Transl Med. 2025 May 28;23(1):597. doi: 10.1186/s12967-025-06404-7.
The multi-kingdom gut microbiota (e.g., bacteriome, mycobiome, and virome) characteristics of colorectal cancer have been extensively studied, yet there is still an insufficient description of the microbiota features in its early-stage, colorectal adenoma, particularly in the gut virome aspect.
Based on the Metagenomic Gut Virus catalogue (MGV) containing 54,118 non-redundant gut viral genomes, this study characterized the virome composition and diversity using publicly available metagenomic sequencing data from 419 individuals with premalignant colorectal adenoma and 552 healthy controls. Furthermore, we identified and assessed the reliability and classification performance of adenoma-associated microbial signatures through comparative analysis and the random forest model.
Our results revealed a notable shift in the gut virome structure of patients compared to healthy controls, characterized by a significant increase in viral families such as Microviridae, Podoviridae_crAss-like, and Quimbyviridae. At the viral operational taxonomic unit (vOTU) level, we identified 479 vOTU signatures showing significant differences in relative abundances between patients and controls, including some patient-enriched vOTUs tending to infect Bacteroidaceae and Lachnospiraceae. Correlation network analysis revealed specific bacterial species correlated with adenoma-associated viruses, suggesting frequent interactions between them. Moreover, random forest models trained on gut viral and bacterial signatures demonstrated area under the curve (AUC) scores of 0.68, 0.82, and 0.76 for classifying healthy individuals versus patients with tubular adenomas, patients with sessile serrated adenomas, and patients with both conditions, respectively. In three independent validation cohorts, the classification performance achieved AUC scores ranging from 0.61 to 0.65.
Our study provides insights into the gut virome in premalignant colorectal adenoma, highlighting its potential role in disease development and diagnosis. Further investigations are warranted to elucidate the underlying mechanisms of gut virus-bacteria interactions and validate diagnostic models in larger populations.
结直肠癌的多界肠道微生物群（如细菌组、真菌组和病毒组）特征已得到广泛研究，但对其早期阶段结直肠腺瘤的微生物群特征，尤其是肠道病毒组方面的描述仍不充分。
基于包含54118个非冗余肠道病毒基因组的宏基因组肠道病毒目录（MGV），本研究利用来自419例结直肠腺瘤癌前病变个体和552例健康对照的公开宏基因组测序数据，对病毒组组成和多样性进行了表征。此外，我们通过比较分析和随机森林模型，鉴定并评估了腺瘤相关微生物特征的可靠性和分类性能。
我们的结果显示，与健康对照相比，患者的肠道病毒组结构发生了显著变化，其特征是微小病毒科、类crAss噬菌体科和昆比病毒科等病毒家族显著增加。在病毒操作分类单元（vOTU）水平上，我们鉴定出479个vOTU特征，其在患者和对照之间的相对丰度存在显著差异，包括一些在患者中富集的vOTU，它们倾向于感染拟杆菌科和毛螺菌科。相关性网络分析揭示了与腺瘤相关病毒相关的特定细菌物种，表明它们之间频繁相互作用。此外，基于肠道病毒和细菌特征训练的随机森林模型在区分健康个体与管状腺瘤患者、无蒂锯齿状腺瘤患者以及同时患有这两种疾病的患者时，曲线下面积（AUC）得分分别为0.68、0.82和0.76。在三个独立的验证队列中，分类性能的AUC得分在0.61至0.65之间。
我们的研究为结直肠腺瘤癌前病变中的肠道病毒组提供了见解，突出了其在疾病发展和诊断中的潜在作用。有必要进一步研究以阐明肠道病毒与细菌相互作用的潜在机制，并在更大规模人群中验证诊断模型。
