3.1 Sequencing and Alpha Diversity Analyses
Alpha diversity refers to the indicators of richness, diversity, and evenness of species in a local uniform habitat. It also knows as within- habitat diversity. In order to comprehensively evaluate the Alpha diversity of microbial communities, Chao1 index is used to characterize richness and Shannon index is used to characterize diversity. The results showed that the species diversity in the samples of the FJ3M, FJ6M, and FJ33M groups was significantly higher than that in the DZ0M and FJ1M groups (P<0.05). However, species uniformity index Simpson had no significant difference among the groups. BA species richness results based on different fermentation time and number of times and compared with MJ, the fermentation environment had an effect on the richness of flora in BA. As shown in table1.
Table 1. Sequencing data and alpha diversity index.
Sample
|
Raw Sequence Number
|
Effective
Sequence Number
|
Goods Coverage
|
ASV/OTU
|
Ace
|
Chao1
|
Shannon
|
Simpson
|
YX1
|
136055
|
128443
|
0.997
|
63
|
1122.6
|
1241.30
|
4.842
|
0.855
|
YX2
|
139155
|
131352
|
0.997
|
92
|
1073.4
|
1178.22
|
4.959
|
0.806
|
YX3
|
127799
|
120231
|
0.997
|
96
|
1269.8
|
1388.77
|
5.531
|
0.839
|
MJ1
|
98553
|
92168
|
0.996
|
67
|
1122.9
|
1209.59
|
5.120
|
0.832
|
MJ2
|
128058
|
120018
|
0.993
|
104
|
1640.7
|
1857.87
|
6.825
|
0.966
|
MJ3
|
126789
|
119672
|
0.995
|
81
|
1304.4
|
1498.65
|
5.177
|
0.824
|
DZ0M1
|
128289
|
122272
|
0.997
|
52
|
758.1
|
827.84
|
4.624
|
0.896
|
DZ0M2
|
139849
|
133292
|
0.998
|
37
|
469
|
557.83
|
3.532
|
0.811
|
DZ0M3
|
154606
|
147303
|
0.996
|
52
|
841.4
|
978.302
|
4.297
|
0.860
|
FJ1M1
|
105724
|
100796
|
0.997
|
27
|
833.3
|
914.613
|
3.843
|
0.826
|
FJ1M2
|
119809
|
113165
|
0.994
|
47
|
1295.6
|
1522.77
|
3.866
|
0.691
|
FJ1M3
|
110109
|
102438
|
0.997
|
23
|
916.9
|
1012.50
|
3.750
|
0.770
|
FJ3M1
|
117508
|
110647
|
0.994
|
60
|
1494.8
|
1651.32
|
5.214
|
0.871
|
FJ3M2
|
113774
|
106864
|
0.995
|
75
|
1611.2
|
1719.77
|
6.407
|
0.953
|
FJ3M3
|
135124
|
127700
|
0.991
|
86
|
1680.9
|
2030.31
|
5.469
|
0.881
|
FJ33M1
|
122109
|
113931
|
0.993
|
75
|
1564.1
|
1757.22
|
5.750
|
0.907
|
FJ33M2
|
98212
|
93070
|
0.997
|
50
|
1044.2
|
1114.14
|
4.141
|
0.695
|
FJ33M3
|
129213
|
123630
|
0.994
|
73
|
1299.9
|
1476.31
|
5.707
|
0.935
|
FJ6M1
|
113151
|
105725
|
0.993
|
73
|
2331.4
|
2451.35
|
6.129
|
0.853
|
FJ6M2
|
98959
|
93594
|
0.997
|
64
|
1352.2
|
1394.99
|
4.922
|
0.845
|
FJ6M3
|
106992
|
100292
|
0.997
|
48
|
2101.6
|
2124.04
|
6.411
|
0.939
|
3.2 Microbial community structure in BA
Through the high-throughput sequencing analysis of 21 samples, we finally got 1846618 tags, and the average tag length was 422.32bp. Finally, cluster analysis was performed by QIIME2 (2019.4) software to obtain an overall range of ASV/OUT of 469-16504, which was used for subsequent species classification analysis. As showed in Figure 1, the number of ASV/OUT increases as the number of samples increases, and the two have a positive correlation. As the sample size increases, the curve tends to be flat. It shows that with the deepening of sequencing, the number of new species obtained will not increase significantly (Rognes, T et al 2016). It can be seen that the amount of samples collected in this project is sufficient and reasonable, which can reflect most of the bacterial information in the samples.
3.2.1 Bacterial composition of BA
The ASV/OTU obtained in the above steps is compared with the data in the NCBI database, and species annotations are made at 6 different species classification levels (phylum, class, order, family, genus, and species). The comparison results showed that ASV/OTU belonged to phylum 31.33, class 58.43, order 48.10, family 266.71, genus 649.10, and species 183.9. Taking into account the characteristics of 16SRNA sequencing, this study mainly described the species composition of the sample flora from the classification level of the phyla and genus. At the phylum classification level, Firmicutes, Proteobacteria, Streptophyta, Bacteroidetes, Actinobacteria were the absolute advantages among the 7 groups of samples Strains.
3.2.2Effect of fermentation time on microbial community structure
As shown in Figure 2 (a), compared with YX, the abundance of Firmicutes, streptophyta, actinobacteria and Bacteroidetes increased in other groups. Based on fermentation time, the abundance of Firmicutes were increased in FJ1M (39.52%), FJ3M (52.82%) and FJ6M (51.70%), The abundance of Actinobacteria were increased in FJ1M (0.62%), FJ3M (5.36%) and FJ6M (12.81%), The abundance of Proteobacteria were not changed very apparent in FJ1M (14.06%), FJ3M (14.15%) and FJ6M (15.63%).
At the genus level, compared with YX, the abundance of Enterococcus, Staphylococcus, Weissella, Leuconostoc were increased in other groups. Based on fermentation time, the abundance of Enterococcus and Flavobacterium were decreased in FJ1M (33.65%, 2.38%), FJ3M (16.15%, 2.83%) and FJ6M (3.40%, 0.20%). The abundance of Staphylococcus was increased in FJ1M (0.78%), FJ3M (2.63%), FJ6M (28.56%). The abundance of Weissella and Lactobacillus were increased first and then decreased in FJ1M (0.23%, 0.79%), FJ3M (17.95%, 3.97%) and FJ6M (3.43%, 1.02%). As shown in Figure 3.
3.2.3Effect of fermentation number of times on microbial community structure
As shown in Figure 2(b), the abundance of Firmicutes were increased in FJ3M (52.92%) and FJ33M (72.90%). The abundance of Streptophyta, Proteobacteria, Bacteroidetes, Actinobacteria were decreased in FJ3M (22.19%, 14.15%, 5.09%, 5.36%) and FJ33M (12.75%, 8.51%, 3.00%, 2.62%). Compared with the bovine bile (DZ0M), The abundance of Firmicutes, Streptophyta, Actinobacteria were increased, while the abundance of Proteobacteria and Bacteroidetes were decreased in Fermentation once (FJ1M, FJ3M, FJ6M) and twice (FJ33M).
At the genus level, the abundance of Weissella, Staphylococcus and Leuconostoc were increased. The abundance of Pseudomonas, Corynebacterium and Dechloromonas were decreased. The abundance of Flavobacterium and Lactobacillus change trend were not obvious. Compared with the bovine bile (DZ0M), The abundance of Lactobacillus, Staphylococcus, Enterococcus and Weissella were increased. The abundance of Flavobacterium, Psychrobacter, Pseudomonas and Acidovorax were decreased. As shown in Figure 2(c).
3.2.4 Beta diversity analyses
To study the similarity of the relationships in the community structure of different samples, Classical Multidimensional Scaling, CMD Scale analysis in R software was used to calculate the distance between the samples. Beta diversity analysis compares the similarity of different samples in species diversity. β diversity uses principal coordinate analysis to visualize all samples in a two-dimensional coordinate system. The similarity distance of the samples is divided into the analysis basis, and the distance between points represents the difference between samples (Yang, G. Z et al. 2021). The two characteristic values of the bacterial community differences among different BA samples were 41.9% (PC1) and 23.9% (PC2). (Figure 4). The results showed that there were significant differences between the BA sample group and the YX group and the DZ0M group, and the flora richness changed significantly. Compared with the FJ6M group, the richness of the FJ1M, FJ3M, and FJ33M groups is different, and the richness of the flora in BA increased with the extension of the fermentation time.
3.2.5Network characteristics of BA
Through the correlation analysis of species abundance information between different samples, the coexistence relationship of species in environmental samples was obtained, and the nature of the DNX microbial network was compared and analysed, highlighting the similarities and differences between the samples (Yi, X et al. 2021). Positive correlation coefficient between species r>0.6, and negative correlation coefficient between species r<0.6. These results show that the network is connected and that the microorganisms in DNX tend to form symbioses. The key flora are Firmicutes, Proteobacteria, Actinobacteria, Bacteroidetes, Streptophyta, Verrucomicrobia. As shown in Figure 4.
3.2.6 Analysis of the Characteristic Bacterial Flora of BA
Linear discriminant analysis (LEfSe) analysis showed that there was a significant difference in the bacterial community between the BA, and emphasizes statistical significance and biological relevance and can identify features with different abundances and associated categories. According to the phylogeny (Figure 5a), can be observed biomarkers of the BA with different fermentation time.
In level LDA score>4 of genus (Figure 5b), Proteobacteria, Bacteria, Flavobacterium, Acidovorax, Rhodospirillales in DZ0M, Eukaryota_Viridiplantae, Streptophyta in FJ1M, Leuconostocaceae, Staphylococcus, Staphylococcaceae, Weissella in FJ33M, Bacillales, Cellvibrio in FJ6M, Bacillus in MJ, Alphaproteobacteria and Liliopsida in YX. The predominant bacteria of BA were different due to the different fermentation times.
3.2.5 Analysis of Metabolic Pathway of BA
Bacterial flora in BA samples were mainly involved in biosynthesis pathways including cofactor, Prosthetic Group, Electron Carrier, and Vitamin Biosynthesis, Fatty Acid and Lipid Biosynthesis, Nucleoside and Nucleotide Biosynthesis, Cell Structure Biosynthesis. The pathways involved in Degradation/ Utilization/ Assimilation include Carbohydrate degradation, Carboxy late Degradation, Organic Nutrient Metabolism, Nucleoside and Nucleotide Degradation, and Secondary Metabolite Degradation. The detoxification pathway mainly involves, Antibiotic Resistance. In Generation of Precursor Metabolite and Energy, it mainly involves fermentation, Glycolysis, TCA cycle, and Pentose Phosphate Pathways. Glycan Pathways mainly involves Glycan Biosynthesis and Glycan Degradation. Metabolic Clusters mainly involve L-glutamate and L-glutamine biosynthesis, O-antigen building blocks biosynthesis (E.Coli), pyrimidine deoxyribonucleotide phosphorylation. (Figure 6).
3.3 Analysis of volatile organic compounds of BA
GC-IMS were used to analyze the volatile components in each sample, DZ0M, FJ1M, FJ3M, FJ6M, FJ33M (comparing blank cattle bile and different stages of fermentation) combined with fingerprint and PCA (Figure 7) diagrams, etc., through the volatile substances can be distinguish by the difference of 4 kinds of samples, but obviously 4 kinds of samples are divided into two groups, FJ1M and FJ3M are one group, and FJ6M and FJ33M are one group.
3.3.1Effect of fermentation time on volatile organic compounds
The substance in area A in the figure shows that the concentration in FJ1M, FJ3M and FJ6M gradually decreases with the fermentation time, mainly: 2-methyl-1-propanol, 3-methyl-1-butanol, ethyl caproate, 3-hydroxy-2-butanone, 2-methyl propanal, octyl aldehyde, E-2-heptanal, nonanal, benzaldehyde ethyl isobutyrate and 2-butanol, etc. The substance in area B in the figure shows that the three samples of FJ1M, FJ3M and FJ6M first increase and then decrease. The concentration in FJ3M is the highest. From left to right, there are mainly: acetic acid, butanol, methyl acetate, and ethyl valerate. Ester, ethyl butyrate, isoamyl acetate, isobutyl acetate, pentanol, propyl acetate, hexanol, ethyl lactate and hexyl acetate, etc. Similarly, in the figure, the substances in the C region (2-butanone, 2-butylfuran, 2-heptanone and 1-penten-3-ol, etc.).
3.3.2Effect of fermentation number of times on volatile organic compounds
The substances in the D region (hexanal and acetaldehyde, etc.) were in the highest concentration of FJ33M. The smell of BA may be caused by these volatile substances. As shown in Figure 8.
3.4 The Link Between Bacteria and the Flavour of BA
Spearman correlation analysis was conducted for the bacteria (top 10) and the main volatile compounds, and the network visualization displayed by Cystoscope software is shown in Figure 9. According to analysis results, Acidovorax, Flavobacterium and Flavobacterium were negatively correlated with Butanol, Isoamyl acetate, 2-Methyl-1-propanol and 3-Methyl-1-butanol, while Staphylococcus, Leuconostoc and Weissella were a positive correlation with them(P<0.05). Enterococcus and Nocardiopsis were a positive correlation with methyl Nonanal, acetate,Ethyl butanoate, Ethyl acetate, Ethyl lactate, 1-Hexanol, Acetic acid, and were negatively correlated with 2-Butanone and acetone(P<0.05). In addition, 3-hydroxy-2-butanone, Isoamyl acetate and isobutyl acetate were a positive correlation with Nocardiopsis, and were negatively correlated with Acidovorax, Flavobacterium, Pseudomonas.
3.5 Metabolic pathways and correlation analysis
The metabolic pathways of volatile components were predicted in the Metaboanalyst database, and the intersections with the fermentation flora metabolic pathway were taken. It was found that the flora in BA may affect through Alanine, aspartate and glutamate metabolism, Arginine biosynthesis, Biosynthesis of amino acids, Butanoate metabolism, Degradation of aromatic compounds, Glyoxylate and decarboxylate metabolism, Methane metabolism, Propanoate metabolism, Pyruvate metabolism Ways affected volatile compounds. (Figure 10).