Determination results of VFAs, NH3-N and CL
There were certain differences for VFAs in the rumen fluid of Tibetan goat of different sex (Table 1). The total VFAs content of rams were extremely higher than that of ewes (P<0.01). Acetic acid was significantly difference between different sexes, which was manifested as that rams were significantly higher than ewes (P<0.05). The concentration of propionic acid, isobutyric acid, butyric acid, isovaleric acid and valeric acid in ewes was higher than that in ewes, but there was no significant difference (P>0.05). The NH3-N content in rams was higher than that of ewes (P>0.05). The fluid cellulase (CL) activity showed that rams were higher than ewes (P>0.05).
Table 1 Rumen fermentation parameters of Tibetan goat of different sex
NOTE:TGM indicates rams,TGFM indicates ewes;A/P indicates Acetic acid/Propionic acid.
Expression determination of VFAs absorption related genes
There were differences in expression of genes related to VFAs transport in rumen epithelium between ewes and rams (Fig. 1). The relative expression levels of AE2, DRA, NHE1, NHE2 and MCT-1 in ewes were significantly higher than those in rams (P<0.05). The expression levels of DRA, MCT-1 and NHE1 were extremely significant difference between ewes and rams (P<0.01) and there was no significant difference in the expression of MCT-4 (P>0.05). Moreover, The expression of MCT-1 in ewes was 1544 times, 16 times, 7538 times, 56 times and 130 times of AE2, DRA, MCT-4, NHE1 and NHE2, respectively.
Diversity of rumen microbiota
A total of 612,007 pairs of reads were obtained in this study, and 610,284 clean reads were generated after tiling and filtering of double-ended reads, and at least 62,730 clean reads were generated for each sample, with an average of 76,286 clean reads were produced. Usearch software was used to cluster Tags at the similarity level of 97% to obtain the number of OTUs of each sample. A total of 969 OTU were obtained, including 945 OTU in rams and 945 OTU in ewes, and 12 unique OTU for each(Fig. 2A). The dilution curve described the species diversity and species richness of each sample, and the flattening of the curve at 30,000 reads indicated that the sequencing coverage was saturated (Fig. 2B). Alpha diversity analysis (Table 2) showed that the ewes were lower than the rams about ACE and Chao1 indexes, but the difference was not significant (P > 0.05), while the Shannon and Simpson indexes were that the ewes were higher than the rams, but the difference was also not significant (P > 0.05).
Table 2 Alpha diversity
The composition of rumen microbiota
At the taxonomic level, a total of 15 phyla, 14 classes, 31 orders, 55 families, 129 genera , and 156 species were detected in this study. At the phylum level, Bacteroidetes and Firmicutes were the dominant phylum, and their relative abundances were all greater than 1% (Fig. 3A). The relative abundance of Bacteroidetes and Firmicutes was the highest in both ewes and rams, accounting for more than 70% of the total abundance. In addition, different species at the phylum level such as Firmicutes, Fibrobacteres and Spirochaetes were the dominant phyla, and the relative abundance of three different species in ewes was significantly higher than that of rams (P<0.05). At the genus level (Fig. 3B), there were 56 genera with relative abundance greater than 0.1%, and Prevotella_1 was the dominant genus in the rumen of both rams and ewes.A total of 13 different species were identified among 129 genera (P<0.05). The relative abundance of Fibrobacter, Ruminococcus_1, uncultured_bacterium_f_Prevotellaceae, Erysipelotrichaceae_UCG-004, Oscillospira, Pyramidobacter and uncultured_bacterium_f_Mitochondria in ewes was significantly higher than that of rams (P<0.05). In addition, the relative abundance of Lachnospira and Ruminococcaceae_NK4A214_group was significantly higher than that of ewes (P<0.05). LEfSe analysis of samples between groups found (Fig. 4) that there were statistically different biomarkers between ewes and rams (LDA score>4), and there were seven differential biomarkers in rams, and differences were significant between ewes and rams.
Prediction of rumen microbiota function
A total of 46 KEGG gene families and 25 COG gene families were identified in the 16S rRNA gene sequencing data using PICRUSt software to predict gene function. Among these predictions, 2 COG gene families showed significant differences between ewes and rams.Of the 43 KEGG gene families, the vast majority were pathways related to metabolism, of which the largest proportion was the function related to carbohydrate metabolism, followed by Amino Acid metabolism and energy metabolism. There were no significant differences that amino acid metabolism and nucleotide metabolism were higher than in ewes that in rams (P>0.05). Among the 25 COG gene families, the functions related to Replication, recombination and repair, RNA processing and modification (Table 3) showed that ewes were significantly higher than rams (P<0.05);
Table 3 Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis
Determination of rumen microbial community density
Fig. 5 shows that there were significant differences in the density of the rumen microbiota of different sex in Tibetan goat. Butyrivibrio fibrisolvens had the highest relative expression in Tibetan goat of different sex, and was significantly higher than other microbiota (P<0.05). In the rumen microbiota of different sex, the relative expression of Fibrobacter succinogenes, Ruminococcus albus and Metha nogenic bacteria were higher in ewes than in rams, while Ruminococcus albus and Methanogens were significantly different between rams and ewes (P<0.05). The relative expression levels of Butyrivibrio fibrolytica, Ruminobacter amylophilus and Ruminococcus flavanum were higher in rams than ewes, and the difference was not significant (P>0.05).
Interactions between rumen microbiota,VFAs and expressions of genes related to transport
As shown in fig. 6, the correlation heat map (correlation threshold>0.5) was constructed between rumen microbiota (top 20 genus level microorganisms with relative abundance) and VFAs and rumen epithelium mRNA expression (2-∆∆CT) of Tibetan goat. There was significant negative correlation between VFAs and the three genus level microorganisms (P<0.05) and significant positive correlation between VFAs and other three genus level microorganisms (P>0.05). Furthermore, Ruminococcaceae_NK4A214_group and Prevotellaceae_UCG-003 showed significant positive correlated with VFAs (P<0.05),
while Ruminococcaceae_NK4A214_group was positively correlated with acetic acid in a significant way (P<0.05). Moreover, Treponema_2, Fibrobacter and Ruminococcus_1 showed significant correlated with VFAs (P<0.05). Uncultured_bacterium_f_Lachnospiraceae, Butyrivibrio_2 and Prevotellaceae_UCG-001 were negatively correlated with ammonia nitrogen (P<0.05).
In addition, the genes related to VFAs transport were correlated with eight genus level microorganisms. Ruminococcus_1 was significantly positively correlated with NHE2 and MCT-1 (P<0.05), and negatively correlated with acetic acid (P<0.05). Fibrobacter showed significant positive correlation with NHEI (P<0.05), and significant negative correlation with acetic acid (P<0.05). Prevotellaceae_UCG-003 was significantly negatively correlated with MCT-1 (P<0.05), and positively correlated with acetic acid (P<0.05).