目的: 通过消化道微生态体外模拟系统探索纳米二氧化钛(titanium dioxide nanoparticles，TiO₂ NPs)对人源肠道菌群组成和结构的影响。方法: 对TiO₂ NPs进行粒径、形状、晶型和团聚程度的表征。通过模拟胃、小肠、结肠的液体环境和物理条件建立体外人消化道微生态模拟系统，并对模拟系统的稳定性进行评价。从人粪便中提取菌群，采用该模拟系统稳定培养，分别暴露于0、20、100、500 mg/L TiO₂ NPs中，收集染毒24 h后的菌液，通过16S rRNA测序技术分析TiO₂ NPs对人源肠道菌群组成和结构的影响，利用线性判别效应量分析(linear discriminant analysis effect size，LEfSe)筛选差异细菌，根据京都基因与基因组百科全书(Kyoto encyclopedia of genes and genomes，KEGG)数据库进行功能预测。结果: TiO₂ NPs为球形粒径，直径(25.12±5.64) nm，晶体结构为锐钛矿，在超纯水中水合粒径为(609.43±60.35) nm，Zeta电位为(-8.33±0.22) mV。体外消化道模拟系统培养人源肠道菌群24 h后达到相对稳定状态，肠球菌(Enterococci)、大肠杆菌(Enterobacterium)和乳酸杆菌(Lactobacillus)计数可分别达到(1.60±0.85)×10⁷个、(5.60±0.82)×10⁷个和(2.70±1.32)×10⁷个。16S rRNA测序结果显示，与对照组相比，TiO₂ NPs染毒组(20、100和500 mg/L)在门、纲、目、科、属水平上，肠道菌群的物种数量和均匀度未受到明显影响，但部分菌种的相对丰度发生显著变化。TiO₂ NPs染毒组(20、100和500 mg/L)与对照组之间共筛选出42种不同的差异细菌(线性判别分析分数，linear discriminant analysis score，LDA score>3)，以肠杆菌属(Enterobacter)、拟杆菌科(Bacteroidaceae)、乳酸杆菌科(Lactobacillaceae)、双歧杆菌科(Bifidobacteriaceae)和梭菌属(Clostridium)等为代表。进一步的肠道菌群功能预测分析显示，TiO₂ NPs可能影响肠道菌群的氧化磷酸化、能量代谢、磷酸盐和磷酸盐代谢、甲烷代谢等代谢和功能(P < 0.05)。结论: 体外人消化道微生态模拟系统下，TiO₂ NPs可以显著改变人源肠道菌群的组成和结构，以肠杆菌属和益生菌为代表，并可能进而影响机体多种物质的代谢和功能。

Objective: To explore the effects of oral exposure to titanium dioxide nanoparticles (TiO₂ NPs) on the composition and structure of human gut microbiota. Methods: The particle size, shape, crystal shape and degree of agglomeration in ultrapure water of TiO₂ NPs were characterized. The in vitro human digestive tract microecological simulation system was established by simulating the fluid environment and physical conditions of stomach, small intestine and colon, and the stability of the simulation system was evaluated. The bacterial communities were extracted from human feces and cultured stably in the simulated system. They were exposed to 0, 20, 100 and 500 mg/L TiO₂ NPs, respectively, and the bacterial fluids were collected after 24 h of exposure. The effect of TiO₂ NPs on the composition and structure of human gut microbiota was analyzed by 16S rRNA sequencing technology. Linear discriminant analysis effect size (LEfSe) was used to screen differential bacteria, and the Kyoto encyclopedia of genes and genomes (KEGG) database for functional prediction. Results: The spherical and anatase TiO₂ NPs were (25.12±5.64) nm in particle size, while in ultra-pure water hydrated particle size was (609.43±60.35) nm and Zeta potential was (-8.33±0.22) mV. The in vitro digestive tract microecology simulation system reached a relatively stable state after 24 hours, and the counts of Enterococci, Enterobacte-rium, and Lactobacillus reached (1.6±0.85)×10⁷, (5.6±0.82)×10⁷ and (2.7±1.32)×10⁷, respectively. 16S rRNA sequencing results showed that compared with the control group, the number and evenness of gut microbiota were not significantly affected at phylum, class, order, family and genus levels in TiO₂ NPs groups (20, 100 and 500 mg/L). The relative abundance of some species was significantly changed, and a total of 42 different bacteria were screened between the TiO₂ NPs groups (20, 100 and 500 mg/L) and the control group [linear discriminant analysis(LDA) score>3], represented by Enterobacter, Bacteroidaceae, Lactobacillaceae, Bifidobacteriaceae and Clostridium. Further predictive analysis of gut microbiota function showed that TiO₂ NPs might affect oxidative phosphorylation, energy meta-bolism, phosphonate and phosphonate metabolism, and methane metabolism (P < 0.05). Conclusion: In human digestive tract microecological simulation system, TiO₂ NPs could significantly change the composition and structure of human gut microbiota, represented by Enterobacter and probiotics, and may further affect a variety of metabolism and function of the body.