A total of seventy (n = 70) 28 day old healthy mice (initial weight 20 ± 2 g) were purchased from the Experimental Animal Center, Zhengzhou University (Zhengzhou, China). The proportion of males and females in each group was 1:1 to decrease the influence of sex on the microbial community. The mice self-propagated through the experimental animal center and possessed a similar genetic background. All the screened mice were subjected to the same immunization program and were determined to be free of other diseases before the experiment. Mice were raised together for 3 days and then randomly divided into seven groups, each containing ten mice including a control group (CON); high dose amoxicillin group (AMXH, 50 mg/kg), medium-dose amoxicillin group (AMXM, 20 mg/kg), low dose amoxicillin group (AMXL 5 mg/kg); high dose andrographolide group (APH, 20 mg/kg), medium-dose andrographolide group (APM, 20 mg/kg) and low dose andrographolide group (APL, 2 mg/kg). All the mice were raised in plastic cages for 14 days with a recommended standard breeding temperature (33–35°C), humidity (53–57%) and illumination time (12 h/12 h light/dark cycle). Moreover, supplies such as feed and water were supplied ad libitum for all groups throughout the entire experiment. Three mice were randomly selected for euthanasia in each group and the gut was excised from abdominal cavity. The separated guts were transferred to a sterilized Kraft paper and knotted with cotton rope to decrease the cross-pollution in the different intestinal segments. The contents in the intermediate sites of cecum were immediately collected and stored at −80°C until further analysis.

DNA from the intestinal samples was extracted using a QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. The quantity of DNA was measured using a NanoDrop 2000 UV spectrophotometer (Thermo Scientific, Waltham, MA, USA). Agarose gel (0.8% w/v) electrophoresis was performed to evaluate the quality and quantity of the extracted DNA.

The amplification of the 16S rDNA target region (V3/V4) utilized PCR with primers (5' to 3')_338F: ACT CCT ACG GGA GGC AGC A and 806R: GGA CTA CHV GGG TWT CTA AT. DNA fragments from the gels were recovered using an AxyPrep DNA Gel Extraction Kit (Axygen-Corning, Glendale, AZ, USA). An FLx800 fluorescent microplate reader (BioTek, Winooski, VT, USA) was used for DNA quantification in conjunction with a Quant-iTPico Green dsDNA Assay Kit (Invitrogen, Waltham, MA, USA). The purified PCR products were used for constructing the sequencing library using Illumina TruSeq (Illumina, San Diego, CA, USA) following the manufacturer's specifications. Prior to sequencing, the sequencing libraries were subjected to fluorescence quantification and quality inspection. The qualified libraries were assembled, diluted and mixed in equal proportions. The final libraries were subjected to high-throughput sequencing using a MiSeq sequencing machine (GENEWIZ, Inc.).

The paired-end sequences achieved from high-throughput sequencing were merged into tags and the quality of raw reads were evaluated using QIIME software (Qiime1.9.1). The reads that passed the initially quality screening were assigned to the corresponding samples according to the primer and barcode information and interrogative sequences such as ambiguous bases, chimeras and mismatched primers were discarded. The obtained high-quality sequences were clustered in a operational taxonomic unit (OTU) on the basis of 97% similarity. Phylogenetic analysis and classifications were performed for the representative sequences for each OTU. The OTU richness distributions were recorded to calculate the alpha diversity. Beta diversity was performed using QIIME (Version 1.7.0) to identify similarities and differences between different samples. In addition, rarefaction curves were constructed to assess the sequencing depth. Linear discriminant analysis effect size (LEfSe) was generated to investigate the differentially abundant taxon. R (v3.0.3) and GraphPad Prism (version 8.0c) were applied to statistical analysis. The criterion of significance was conducted at p-values < 0.05 and the data was expressed as means ± SD.

We initially performed a quality screening of our high throughput sequencing data to eliminate erroneous and questionable sequences to verify sequence reliability. The data was optimized and produced 1,756,513 high-quality reads with an average of 56,662 reads (range 53,560–106,150) per sample (Table 1) and a sequence length of 443–460 bp (Figure 1).

The qualified reads comprised 110 OTUs on the basis of 97% nucleotide-sequence identity (Figure 2A). The rarefaction and rank abundance curves per sample were relatively flat and displayed a tendency to saturate suggesting that the depth and evenness of the sequences meets the requirements for sequencing and analysis (Figures 2B–D).

We assessed the alpha diversity of our gut microbiota samples and the Good's coverage estimates varied from 99.8 to 99.9% for all of the samples indicating excellent coverage (Table 2). The Chao1 indices for the experimental groups AMXL, AMXM, and AMXH were 354.61, 323.79, and 250.54 and the corresponding ACE indices were 355.85, 327.99, and 249.51, respectively. Moreover, the averages of Shannon indices for these 3 groups were 4.67, 4.44, and 4.27, respectively. The Chao1, ACE, Shannon, and Simpson indices for these groups treated with AMX displayed a gradual downward trend as the drug concentration was increased indicating that AMX reduced the abundance and diversity of the intestinal microbial community. Remarkably, the three diversity indices (ACE, Chao1, and Shannon) of the control group were higher than those of the AMX treatment groups. In contrast, significant differences in gut microbiota abundance and diversity were only observed between the control and AMXH group. The average for the Chao1 index in the andrographolide treatment groups (APH, APM and APL) ranged from 412.64 to 463.67 while the ACE index ranged from 405.47 to 439.24. Moreover, the average values for the Shannon index in the andrographolide treatment group ranged from 5.50 to 5.70. Interestingly, the Chao1 and ACE index for APH was higher compared with APL and APM and the 4 diversity indices for APM were lower when compared to APL, APM and CON. However, differences between the 4 groups were not significant (P > 0.05). The α-diversity indices revealed no significant difference in the diversity and richness of gut microbiota between the andrographolide and negative control groups. Both the weighted and the unweighted PCoA plots revealed that the samples in each group were clustered separately indicating differences in the gut microbiota for the samples (Figure 3).

We obtained that were comprised of 11 phyla, 17 classes, 28 orders, 55 families, 157 genera and 168 species. The Firmicutes, Bacteroidetes and Proteobacteria were the 3 dominant phyla for all samples and represented 45.09, 40.46, and 11.12%, of the totals, respectively. These phyla constituted the core of the microbiota and accounted for 96.67% of the taxonomic groups identified. The andrographolide treatment groups were primarily composed of Firmicutes (66.21%) and Bacteroidetes (27.51%) and Proteobacteria (1.83%). The Firmicutes, Bacteroidetes and Proteobacteria content for the AMX treatment groups accounted for 24.88, 50.07, and 22.97% of the total, respectively. The dominant phyla for the negative control group were Bacteroidetes, Firmicutes, and Proteobacteria and were accounted for by 50.55, 40.40, and 3.46% of the totals, respectively (Figure 4).

The genus Lachnospiraceae_NK4A136 comprised 24.64 and 21.65% of the overall bacterial composition in the APH and APM groups while f_Lachnospiraceae_Unclassified were represented at 16.74 and 15.80%, respectively. The levels of these two genera in the APH and APM groups were significantly higher than the negative control and AMX treatment groups (P < 0.05). In the APL test group, the most numerous genera were f_Muribaculaceae Unclassified and Lachnospiraceae_NK4A136 at 22.79 and 19.77% of the overall bacterial composition, respectively. The levels of the latter genra were significantly higher than the negative control and AMX groups (P < 0.05).

The most abundant genera for the AMX treatment groups were Bacteroides at 35.87, 41.11, and 30.26% for AMXL, AMXM, and AMXH, respectively. Klebsiella was second most abundant for these groups at 13.74, 13.94, and 8.45%, respectively. These levels for Bacteroides and Klebsiella were significantly higher than for the negative control and andrographolide treatment groups (Figure 5).

We performed a metastatic analysis to identify significantly different taxa between the different test groups (Tables 3–11). Our results revealed that at the genus level, the abundance of Candidatus_Saccharimonas increased while the Lactobacillus, Escherichia-Shigella and Alistipes genera decreased significantly in the APL group; The abundance of Ruminococcaceae also increased significantly while Alistipes decreased significantly in the APM and APH groups, respectively. The abundance of Bacteroides in the AMXL and AMXM groups was higher than negative controls while Lactobacillus levels were lower than negative controls. In the latter, Corynebacterium_1 and Klebsiella increased significantly while f_Muribaculaceae_Unclassified and Alipis decreased significantly in the AMXH group. Compared with the AMX test groups, Lachnospiraceae_NK4A136p, Lactobacillus and Candidatus_Saccharimonas abundance in the APL group were significantly higher than for the AMXL group while Bacteroides, Escherichia-Shigella, and Phoscolarctobacterium in the APL group were significantly lower than for the AMXL group. Compared with the AMXM group, the abundance of Lachnospiraceae_NK4A136, Ruminococcus_1, f_Ruminococcaceae_Unclassified and [Eubacterium]_xylanophilum_group increased significantly and was accompanied by a significant decrease in the abundance of Bacteroides Escherichia-Shigella, Blautia and Parasutterella for the APM group. The AMXH and APH groups displayed a significant increase in the abundance of f_Lachnospiraceae_Unclassified, Lachnospiraceae NK4A136_group, Ruminococcaceae_UCG-014, Candidatus_Saccharimonas and Roseburia as well as a significant decrease in the abundance of Phascolarctobacterium and Corynebacterium_1.