Effect of the diet level of whole-plant corn silage on the colonic microflora of Hezuo pigs

Experimental design

WPCS was provided by a dairy plant in Yongdeng Farm (Gansu Province, China). The WPCS was naturally air-dried and crushed into powder to prepare experimental diets, and the traditional nutritional levels of WPCS are shown in Table 1.

The experimental rations were formulated according to the China feeding standard of swine (NY/T 65-2004, https://www.antpedia.com/standard/5041407.html), the crushed WPCS powder was added to replace the basal diet at the proportions of 5%, 10% and 15% to total weight, and ensured that meet the pig energy requirements. Once the test diets had been prepared and mixed uniformly, they were placed in ration pouches with inner bags to prevent moisture ingress. The dietary ingredients of experimental pigs and the nutrient contents of experimental diets are shown in Tables 2 and 3, respectively.

Forty healthy Hezuo pigs at 60 days of age with similar body weight were selected and fed with the basal diet in the pre-feeding period for 7 days. Then, 32 Hezuo pigs, which had no differ in terms of feed intake and weight gain (body weight 7.88 ± 0.81 kg), were selected and divided equally into four treatment groups, with eight replicates per group (with half of male and half of female) and one pig per replicate. During the main experimental period, the control group was fed the basal diet, and the experimental groups I, II, and III were fed experimental diets that was replaced with 5%, 10%, and 15% WPCS in basal diet, respectively, for 120 days. The feed amount was based on 3% of live weight, and weighed before and after of each feeding to calculated the feed consumption of each pig. All pigs were raised in the same house and kept under the same housing and management conditions, and each replicate was housed in one single pen for the duration of the experiment. They were fed three times a day at 08:00, 13:00, and 19:00, and were given ad libitum access to water. Vaccinations, deworming, cleaning and disinfection are in line with routine pig farm practice. The average of ambient temperature and relative humidity of the pig housing were recorded as (24 ± 2) °C and 60% to 65%, respectively.

The experiment was conducted from June to October 2022 at the Gansu Bailu Eco-Tourism Farm (Lanzhou, China). All experimental protocols used in this study were reviewed and approved by the Animal Protection and Use Committee of Gansu Agricultural University (approval number: GSAU-Eth-AST 2022-024).

Sample collection and sequencing procedures

At the end of the experiment, six pigs in each group were randomly selected for sacrifice after barbiturate anesthesia. The intestines were separated and the contents samples in colonic midpiece were collected in 5 mL sterilized centrifuge tubes, which were immediately put into liquid nitrogen quick-freezing, and stored at −80 °C.

Illumina sequencing of the 16S rRNA gene was carried out to characterize the diversity and composition of the microbial community.

Sequencing analysis of colonic content samples for bacterial flora

A TIANGEN DP302-02 bacterial genomic DNA extraction kit (TIANGEN, Beijing, China) was used to extract total bacterial DNA, following the instructions provided. The extracted total DNA was purified and its purity and concentration were determined by 2% agarose gel electrophoresis, and the samples were diluted to 1 ng/μL with sterile water. Diluted bacterial genomic DNA was used as a template for amplification the V4 hypervariable region of the 16S rRNA gene using the primers (515F: 5′-GTGCCAGCMGCCGCGGGTAA-3′ and 806R: 5′-GGACTACHVGGGTWTCTAAT-3′). The library was constructed using the NEB Next^® Ultra^™ II FS DNA PCR-free Library Prep Kit (New England Biolabs, Ipswich, MA, USA) following the manufacturer’s recommendations. The constructed library was quantified by Qubit and qPCR. After library qualification, the NovaSeq 6000 was used for on-line sequencing of PE250 at Novogene Bioinformatics Technology Co.

Bioinformatics and statistical analysis

The acquired reads were merged using FLASH (Version 1.2.11, http://ccb.jhu.edu/software/FLASH/) (Magoc & Salzberg, 2011) and quality filtering of raw reads was strictly filtered by FASTP software (Version 0.23.1) to obtain high quality clean tags following the QIIME (V1.9.1) quality control procedure (Bokulich et al., 2013; Caporaso et al., 2010). Subsequently, the clean tags were compared with the species annotation database (Silva database, https://www.arb-silva.de/) to detect chimeric sequences. The chimeric sequences were then removed to obtain the final effective tags (Edgar et al., 2011).

The Uparse algorithm (Uparse v7.0.1001) is used to cluster all effective tags (Edgar, 2013), and the sequences were clustered into OTUs (Operational Taxonomic Units) with a 97% identity threshold. The most abundant sequences in the OTUs were screened as representative sequences for annotating species (Silva 138.1 database, https://www.arb-silva.de/documentation/release-1381/) (Quast et al., 2013). All raw sequences of the 16S rRNA gene have been deposited in the Sequence Reads Archive (SRA) under the BioProject accession number PRJNA1071212.

Venn diagrams are created in using the VennDiagram function in R software (version 4.0.3). The community alpha diversity was calculated using Qiime2 (Version 2023.5). Principal coordinates analysis (PCoA) analyses were calculated and plotted using the ade4 and ggplot2 packages in R software package (version 4.0.3; R Core Team, 2020). Based on the species annotation results, the top 10 species in terms of relative abundance were selected to plot the histogram of species abundance for each group at phylum and genus levels using the SVG function in Perl. The T-test in R was used to analyze the significant differences of species. LEfSe analysis was performed using Lefse software with the default setting of LDA score threshold of 4. The function of microflora was predicted using PICRUSt2 (V2.3.0) software.

The statistical differences in alpha diversity of bacterial communities were detected using Tukey test. Differentially abundant bacterial taxa and bacterial function between the control group and the three experimental groups were detected by T-test. The differential analyses of bacterial genera of control vs group I, control vs group II, and control vs group III was conducted using the DESeq2 package in RStudio (version 1.20.0; R Studio Team, 2020). Genera with padj ≤ 0.05 and |log₂FoldChange| ≥ 1 were classified as either upregulated or downregulated across the various dietary groups.

Effect of whole-plant corn silage on the growth performance of Hezuo pigs

In our prior research, we evaluated the impact of varying levels of whole-plant corn silage incorporation on the growth performance metrics, including body weight, average daily feed intake, and feed conversion ratio in Hezuo pigs (Yin et al., 2024; Table 4), and found that the feed conversion ratio in the 10% supplementation group (Group II) was substantially lower than that of the control group (P < 0.05), while the other groups exhibited no substantial differences (P > 0.05). The average daily feed intake and average daily weight gain showed no significant variation across the four groups (P > 0.05).

Venn diagram of the colonic microflora in Hezuo pigs

As shown in Fig. 1, a total of 5,655 OTUs were obtained from the colonic content samples of all experimental pigs. Of these, 979 OTUs were common to the four groups, while 771, 691, 1,575, and 471 OTUs were specific to the control group, group I, group II, and group III of experimental pigs, respectively. Notably, the pigs in experimental II exhibited the highest species richness of colonic flora among the four groups.

Alpha diversity of the colonic microflora in Hezuo pigs

Figure 2 demonstrated that there was no significant difference for the Chao1 and Observed_species indexes among the control, group I, and group II (P > 0.05), while these indexes were all significantly higher than those observed in group III (P < 0.05, Figs. 2A and 2B). The Shannon and Simpson indexes were found to be significantly higher in group I and II in comparison to the control group, and significantly lower in group III in comparison to the control group, group I and group II (P < 0.05, Figs. 2C and 2D).

PCoA analysis of the colonic microflora in Hezuo pigs

As illustrated in Fig. 3, principal components PC1 and PC2 accounted for 15.22% and 13% of the total variables, respectively. This indicates good clustering of samples in each group. The colonic microflora of the control group and test group I Hezuo pigs were more closely clustered, and the samples of test group II and test group III were more closely clustered. What’s more, the colonic microflora in control and group I of Hezuo pigs makes sharp distinctions to the groups II and III, indicating that adding 10% and 15% WPCS in the ration of Hezuo pigs have resulted in significant changes of the structure and diversity of the colonic microflora.

Composition of the colonic microflora in Hezuo pigs

Figure 4 presents the most abundant top 10 species at the phylum and genus level for each group of Hezuo pig. The predominant bacterial phyla in the colonic microflora of Hezuo pigs were Firmicutes, Bacteroidota, and Euryarchaeota, collectively accounting for over 95% of the colonic microflora in both the control and group II. The remaining bacterial phyla were Spirochaetota, Proteobacteria, Chloroflexi, Fibrobacterota, Desulfobacteroa, Verrucomicrobiota, and Actinobacteriota (Fig. 4A). As shown in Fig. 4B, the most prevalent bacterial genera in the colon of Hezuo pig were Clostridium sensu stricto 1 and Rikenellaceae RC9 gut group with the highest abundance observed in the four groups. Additionally, Streptococcus was more prevalent in the control group, p-251-o5 was more common in the three experiment groups. The remaining major bacterial genera were Parabacteroides, Methanobrevibacter, Bacteroides, Lactobacillus, and Prevotella.

Analysis of the differences of the colonic microflora in Hezuo pigs

The differences of the colonic bacterial genera in Hezuo pigs between the control group and the three experimental groups are presented using a volcano plot and T test bar plot, respectively (Figs. 5 and 6). In comparison to the control group, there were 14, and five upregulated genera, five, and 10 downregulated genera in groups I, and II, respectively (Figs. 5A and 5B). The group III is the most exclusive compartment, enriching for 20 genera while downregulating 24 genera (Fig. 6C). Relative abundance of Streptococcus was significantly decreased in experimental groups I, II, and III (P < 0.05, Fig. 6). Relative abundances of Lachnospira, Eubacterium_xylanophilum_group and Eubacterium_siraeum_group in experimental groups II and III were observed significantly decreased when comparing with the control group (P < 0.05, Fig. 6). Relative abundances of p-251-o5, Parabacteroides, Prevotellaceae UCG-003, Prevotellaceae UCG-001, and F082 were higher in at least two experimental groups than that in the control group (P < 0.05, Fig. 6). Furthermore, Fibrobacter, Rikenellaceae RC9 gut group in group I, and UCG-010 in group II, as well as the Bacteroides in group III exhibited a significant increased relative abundance as compared with the control group (P < 0.05, Fig. 6).

LEfSe analysis of the differential colonic microflora in Hezuo pigs

The differential colonic microbes among the four groups Hezuo pigs using LEfSe analysis with the standard of LDA score > 4 was shown in Fig. 7. It can be seen that Streptococcus and Lactobacillus were the differential microbial taxa in the control group Hezuo pigs, Prevotellaceae, Lachnospiraceae, Spirochaetales, Treponema and Fibrobacter (and Fibrobacter intestinalis) were the differential microbes in the colon of Group I Hezuo pigs. The differential microflora in the colon of Group II Hezuo pig were Oscillospiraceae, Rikenellaceae_RC9_gut_group, Tannerellaceae, UCG_005 and Christensenellaceae. The differential microflora in the colon of Group III Hezuo pig were p_251_o5, Fibrobacter and Bacteroides.

Microbial functions in the colon of Hezuo pigs

The top 35 microbial functional categories ranked by their relative abundance, were selected for comparison their differences between the control group and the three experimental groups at the secondary functional level of KEGG pathways. The results are shown in Fig. 8. When compared to the control group, cellular process and signaling were significantly increased, and infectious diseases were significantly decreased in all there comparation groups (P < 0.05); cancers, nervous system, immune system and environmental adaptation were all differed from groups I and II (P < 0.05); xenobiotics biodegradation and metabolism was only differed from the group I (P < 0.05). Notably, three predominant pathways of translation, nucleotide metabolism and signal were only differed from the group II (P < 0.05).