DNA extraction and PCR amplification
Fecal samples from different seasons were homogenized to obtain four
groups of mixed samples for spring and autumn (Zhou et al., 2020; Xiao
et al., 2019; Gong et al., 2021). Extraction of total microbial DNA from
fecal samples was performed. The V3-V4 hypervariable regions of the
microbial 16S rRNA gene were amplified by PCR using primers 338F
(5′-ACTCCTACGGGAGGCAGCA-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). The
PCR conditions were as follows: 98°C for2 min; 30 cycles at 98°C for
15s, 55°C for 30s, and 72°C for 30s; and extension at 72°C for 5 min
(Zhou et al., 2020). The products were purified, quantified, and
normalized to generate a sequencing library (SMRT Bell). The constructed
library was first subjected to library quality inspection, and the
qualified products were sequenced using PacBio Sequel at Biomarker
Technologies Co., Ltd. (Beijing, China).
Raw subreads were calibrated to get circular consensus sequencing (CCS)
(SMRT Link, version 8.0), then Lima v1.7.0 software was used to identify
CCS sequences of different samples through barcode sequences and remove
chimeras to obtain Effective-CCS sequences. Operational taxonomic units
(OTUs) were clustered with a 97% similarity cutoff using UPARSE
(version 10.0) (Edgar, 2013).
Alpha diversity index analysis was performed using QIIME2 software, via
the Wilcoxon rank-sum test to compare the community diversity indices
(Ace richness estimator and Shannon-Wiener index) (Prehn-Kristensen et
al., 2018; Schloss et al., 2011).
Dimensionality reduction of data was based on the Bray-Curtis distance
matrix by principal coordinate analysis (PCoA), and nonmetric
multidimensional scaling (NMDS) was used to observe the differences in
gut microbial community structure (Oksanen et al., 2013). We modeled the
binary Jaccard distance dissimilarities based on an operational
taxonomic unit (OTU) level table via analysis of similarities (ANOSIM)
(Anderson, 2001).
Using SILVA as the reference database, the community species composition
analysis of the two groups of experiments was performed at various
levels: phylum, order, family, genus, and species. Linear discriminant
effect size (LEfSe) analysis was used to screen microorganisms with
large differences as potential markers (Segata et al., 2011), the
significance of different species was determined using Matestats
software, and differential strains were screened according to the
criteria of LDA>4, q<0.05. Picrust2 was used to
predict changes in KEGG-enriched functional pathways across the groups
(Kanehisa, 2019).