联合抗生素对伪无菌模型大鼠的肠道菌群及其代谢的影响

The effect of combined antibiotics on the intestinal flora and its metabolism in the construction of a pseudo germ-free rat model

摘要:

目的 比较联合抗生素经灌胃与自由饮水两种给药复制大鼠伪无菌模型的效果，为优化伪无菌模型复制方法提供参考。方法 将SD大鼠随机分为正常组、抗生素灌胃组和抗生素饮水组，每组5只。抗生素灌胃组采用含200 mg/mL氨苄西林、200 mg/mL甲硝唑、200 mg/mL新霉素和100 mg/mL万古霉素的混悬液，按1 mL/kg体积每日灌胃；抗生素饮水组在灭菌饮用水中加入0.5 g/L氨苄西林、0.5 g/L甲硝唑、0.5 g/L新霉素和0.25 g/L万古霉素，大鼠自由饮水，干预14 d。干预后第5、10、14天，无菌取大鼠粪便进行菌群培养；14 d后取大鼠粪便进行16S核糖体rDNA（16S rDNA）测序；分离肝脏、脾脏、肾脏、胸腺称重，取各组大鼠结肠组织进行苏木精-伊红（HE）染色及免疫组织化学染色。结果 与正常组比较，在有氧和厌氧培养条件下，第5天起抗生素灌胃组和抗生素饮水组粪便菌落显著减少，伪无菌大鼠模型复制成功并稳定维持到第14天。与正常组比较，抗生素饮水组第2～6天体重均下降（P <0.05），第7～14天体重差异均无统计学意义（P >0.05）；抗生素灌胃组第6天体重高于正常组（P <0.05）。正常组、抗生素灌胃组和抗生素饮水组大鼠肝脏、肾脏、脾脏和胸腺系数的比较，差异均无统计学意义（P >0.05）。各组大鼠结肠免疫组织化学结果显示，与正常组比较，抗生素灌胃组与抗生素饮水组中紧密连接蛋白闭锁小带蛋白-1（ZO-1）、闭合蛋白（Occludin）在肠上皮和腺上皮细胞细胞质及细胞膜呈阳性表达，棕黄色染色较多。基于16S rDNA测序的α多样性分析显示：与正常组比较，抗生素灌胃组和抗生素饮水组的ACE、Sobs及Shannon指数均降低（P <0.05），Simpson指数均升高（P <0.05）；抗生素饮水组的Simpson指数较抗生素灌胃组升高（P <0.05）。基于Bray-Curtis距离的β多样性分析显示：正常组样本与抗生素灌胃组、抗生素饮水组样本明显分离（ANOSIM相似性分析：r =0.816，P =0.001）。抗生素灌胃组和抗生素饮水组变形菌门丰度均高于正常组（均P <0.05），抗生素灌胃组与抗生素饮水组比较，差异无统计学意义（P >0.05）；正常组厚壁菌门、拟杆菌门及放线菌门丰度高于抗生素灌胃组和抗生素饮水组（均P <0.05）。抗生素饮水组摩根菌属丰度高于抗生素灌胃组和正常组（均P <0.05）；抗生素灌胃组埃希-志贺菌属丰度高于抗生素饮水组和正常组（P < 0.05）；正常组乳酸杆菌属、拟杆菌属及norank_f__Muribaculaceae属丰度高于抗生素灌胃组和抗生素饮水组（均P <0.05）。基于KEGG的菌群功能预测分析结果显示：在碳水化合物代谢、氨基酸代谢、能量代谢、萜类化合物和聚酮类化合物代谢、辅助因子和维生素代谢等多个途径上，3组预测丰度的总体比较，差异均有统计学意义（P <0.05）。结论 联合抗生素经灌胃和自由饮水给药均可在14 d内有效复制SD大鼠伪无菌模型，但抗生素灌胃组与抗生素饮水组均出现条件致病菌异常繁殖现象，这提示抗生素干预可能通过筛选压力改变菌群结构，使原本处于低丰度的潜在致病菌获得生长优势。这为研究者根据实验目标菌群选择适宜的抗生素给药策略复制伪无菌模型提供了参考。

Abstract:

Objective To compare the effects of combined antibiotics administered via gavage and free drinking water to establish a pseudo germ-free (PGF) model in SD rats, and to provide a reference for the optimisation of the construction method of the PGF model.Methods SD rats were randomly divided into 3 groups of 5 rats each: normal group (Group N), antibiotic gavage group (Group I) and antibiotic drinking water group (Group D). Group I was treated with a suspension containing 200 mg/mL ampicillin, 200 mg/mL metronidazole, 200 mg/mL neomycin, and 100 mg/mL vancomycin by daily gavage at a rate of 1 mL/kg volume; In group D, 0.5 g/L ampicillin, 0.5 g/L metronidazole, 0.5 g/L neomycin and 0.25 g/L vancomycin were added to sterilised drinking water and rats were given free access to water for 14 d of the intervention. On the 5th, 10th and 14th days after the intervention, feces of the rats were aseptically taken for bacterial colony culture; after 14 days, rat faeces were taken for 16S ribosomal rDNA (16S rDNA) sequencing; liver, spleen, kidney and thymus were isolated and weighed, and colon tissue was taken for HE and immunohistochemical staining.Results Compared with group N, under aerobic and anaerobic culture conditions, the fecal colonies in groups I and D significantly decreased from day 5. The PGF rat model was established and stably maintained until day 14. However, compared with group N, the body weight of group D significantly decreased from day 2 to day 6 (both P < 0.05), and the difference in weight from day 7 to 14 was not statistically significant (P > 0.05); the weight of the group I was higher than that of the group N on day 6 (P < 0.05). Comparison of liver, kidney, spleen and thymus coefficients of rats in group N, group I and group D showed no statistically significant difference (P > 0.05). Alpha diversity analysis based on 16S rDNA sequencing showed that ACE, Sobs and Shannon indices were decreased (P < 0.05) and Simpson indices were increased (P < 0.05) in the group I and group D compared with the group N; Simpson indices in the group D were increased (P < 0.05) compared with that in the group I. Beta diversity analysis based on Bray-Curtis distance showed that the samples from the normal group were significantly separated from the samples from the group I and group D (ANOSIM similarity analysis: r = 0.816, P = 0.001); Samples from the group I and group D were clustered closer together, indicating a higher degree of similarity in bacterial flora composition between the two groups. The abundance of Proteobacteria was higher in the group I and group D than in the group N (both P < 0.05), and the difference between the group I and group D was not statistically significant (P > 0.05); the abundance of Firmicutes, Bacteroidetes and Actinomycetes was higher in the normal group than in the group I and group D (all P < 0.05). The abundance of Morganella spp. was higher in the group D than in the group I and the group N (both P < 0.05); the abundance of Escherichia-Shigella spp. was higher in the group I than in the group D and the group N (both P < 0.05); the abundance of Lactobacillus spp., Bacteroides spp. and norank_f__Muribaculaceae spp. in the group N was higher than in the group I and group D (both P < 0.05). KEGG-based functional prediction analyses of the bacterial colonies showed statistically significant differences (P < 0.05) in the overall comparison of predicted abundance among the three groups in a number of pathways, including carbohydrate metabolism, amino acid metabolism, energy metabolism, terpene and polyketide metabolism, cofactors, and vitamin metabolism.Conclusion Combined antibiotics administered by gavage and free drinking water were effective in replicating the PGF rat in SD rats within 14 d. However, abnormal multiplication of conditionally pathogenic bacteria occurred in both the group I and the group D, which suggests that antibiotic interventions may change the colony structure through screening pressure, so that potentially pathogenic bacteria, which would otherwise be at a low abundance, may gain a growth advantage. This provides a reference for researchers to replicate the PGF rat model by selecting appropriate antibiotic administration strategies according to the experimental target flora.