Identification of metabolites and screening for differential metabolites
Extracellular and intracellular metabolites were respectively measured by GC-MS on the basis of mass spectrum and retention index match, of which the components of amino acid metabolism, glycolysis, the citric acid (TCA) cycle, organic acids, fatty acids, sugars, and sugar alcohols were identified. A of total 380 peaks and about 157 identified metabolites were detected in fermentation broth samples. A of total 440 peaks and about 198 identified metabolites were detected in mycelial samples. The amount of metabolites in mycelia was more than that of fermentation broth, with the common 95 metabolites (Fig. 1). The 62 and 103 metabolites were respectively limited to fermentation broth and mycelia of Epichloë sp. .
Principal component analysis of Epichloë sp. from F. sinensis
The difference between treated samples from fermentation broth of Epichloë sp. at different time points was assessed using principal component analysis (PCA) (Fig. 2). The first principal component produced the greatest contribution (75.79%), and the second principal component accounted for 9.84% of the total. Using this method, a clear distinction was noted between the samples under different culture conditions (cultivation time and Se concentration).
The processing of the metabolomic data with PCA revealed clustering of mycelial samples according to Se concentration and cultivation time (Fig. 3). The first and second principal components produced accounted for 59.25% and 9.03% of the total sample variance, respectively, indicating the obvious difference between the samples under different culture conditions (cultivation time and Se concentration).
We carried out PCA for visualizing the difference among these metabolites of the fermentation broth (Fig. 4). Analysis of 157 metabolites showed that three principal components, the three components representing about 59.78% of the total variance. Metabolites contributing to the first component were lactic acid, alanine, O-methylthreonine, valine, isoleucine, glycine, serine, L-allothreonine, aminomalonic acid, 4-aminobutyric acid, phenyalanine, 5-aminovaleric acid, 4-hydroxybenzoic acid, cis-1,2-dihydronaphthalene, O-phosphorylethanolamine, ornithine, and tagatose. Metabolites contributing to the second component were acetol, glutamine, 1-methylhydantoin, and citraconic aciddegr. Metabolites contributing to the third component were lyxose, diglycerol, and glucose-1-phosphate.
For visualizing the difference among metabolites of mycelia, two primary components were identified using PCA in an analysis of 198 metabolites, with the two components accounting for approximately 39.74% of the total variance (Fig. 5). Metabolites contributing to the first component included pyruvic acid, alanine, sulfuric acid, malonic acid, valine, hydroxyurea, ethanolamine, glycerol, isoleucine, proline, glycine, succinic acid, uracil, fumaric acid, serine, threonine, β-alanine, aminomalonic acid, aspartic acid, oxoproline, 4-aminobutyric acid, glutamic acid, and phenylalanine. Metabolites contributing to the second component included 4-hydroxy-3-methoxybenzyl alcohol, indole-3-acetamide, and dioctyl phthalate.
The eight common contributed metabolites (alanine, valine, isoleucine, glycine, serine, aminomalonic acid, 4-aminobutyric acid, and phenylalanine) were found in both fermentation broth and mycelia (Fig. 4, 5).
Heatmap and hierarchical cluster analysis of Epichloë sp. from F. sinensis
In order to identify the different metabolites between fermentation broth samples in response to Se concentrations, all metabolite profiles, consisting of 15 Epichloë sp. samples were performed by heatmap and hierarchical cluster analysis (Fig. 6). According to hierarchical clustering analysis of the 157 identified extracellular metabolites, an obvious separation was observed between samples, with 15 samples being clearly grouped into three categories: the first category included two samples (the control samples for weeks 4 and 5 of cultivation), the second category included four samples (both the control and 0.2 mmol/L Na2SeO3 treated samples for week 7 of cultivation, as well as the Se-treated samples in the fourth week), and the third category included the remaining nine samples. Additionally, 157 metabolites were divided into two clusters, consisted of 74 and 83 compounds, respectively.
According to hierarchical clustering analysis of the 198 identified intracellular metabolites, an obvious separation was also noted between samples, with 15 samples being clearly grouped into two classes, one category consisted of 5 samples (both the control and treated samples for week 4, as well as control samples between weeks 7 and 8), and the other consisted of the rest 10 samples (Fig. 7) (both the control and treated samples between weeks 5 and 6, as well as treated samples between weeks 7 and 8). Besides, 198 metabolites were divided into two groups, consisted of 61 and 137 compounds, respectively.
Changes in the extracellular and intracellular metabolite profiles of Epichloë sp. from F. sinensis in response to Se
The changes in identified metabolites of fermentation broth were presented in Tables 1 and 2. Among the 157 identified metabolites, 22 metabolites displayed differential accumulation in the 0.1 mmol/L Se group compared to the CK group (0.1 mmol/L vs. CK), while seventeen metabolites showed decreased levels in the 0.1 mmol/L Se treatment group. We observed an increase in 29 metabolites and a decrease in 13 metabolites in 0.2 mmol/L vs. CK. These findings indicated that Na2SeO3 concentrations affected metabolite alterations. Parallel to the cultivation period, 67 metabolites (ཞ43%) were found to increase while 70 metabolites (ཞ45%) were found to decrease in the fourth week in the Se group compared to the control group (Se vs. CK). In week 5 of cultivation, 87 up-regulated and 58 down-regulated metabolites were detected in Se vs. CK. In week 6, 64 metabolites were up-regulated and 38 ones were down-regulated between Se and control. 77 compounds showed a high accumulation and 43 compounds were suppressed in Se vs. CK during 7 week of cultivation. In the last week of cultivation, 76 compounds showed accumulation at Se concentrations and 54 others decreased. The extracellular samples displayed metabolic differences in terms of cultivation time and Se treatments.
Table 1
Number of increased or decreased metabolite levels among 157 identified metabolites in Epichloë sp. fermentation broth from different Se concentration and control groups.
Change
|
Number of metabolites
|
0.1 mmol/L vs. CK
|
0.2 mmol/L vs. CK
|
Increase
|
22
|
29
|
Decrease
|
17
|
13
|
Table 2
Changes in metabolite profiles in Epichloë sp. fermentation broth between Se and control groups at weeks 4∼8.
Change in
Se vs. CK
|
Number of metabolites
|
week 4
|
week 5
|
week 6
|
week 7
|
week 8
|
Increase
|
67
|
87
|
64
|
77
|
76
|
Decrease
|
70
|
58
|
38
|
43
|
54
|
Changes in metabolites from mycelia of Epichloë sp. between Se and control groups were showed in Tables 3 and 4 by comparing. In 0.1 mmol/L vs. CK, three metabolites increased and 35 decreased, whereas two metabolites increased and 29 decreased in 0.2 mmol/L vs 0 mmol/L samples. 43 metabolites were up-regulated and 83 ones were down-regulated in week 4 of cultivation in the Se group compared with control group. In week 5, 21 accumulations and 98 reductions were detected in Se vs. CK. In the 6th week, 22 metabolites were up-regulated and 90 ones were down-regulated between Se and control groups. In 7 week of cultivation, 24 up-regulated and 109 down-regulated metabolites were obtained in Se vs. CK. During the last week of cultivation, 30 metabolites showed a high accumulation and 91 ones were reduced between Se and control groups. The intracellular samples showed metabolic complexity with regard to cultivation time and Se treatments.
Table 3
Number of increased or decreased metabolite levels among 198 identified metabolites in Epichloë sp. mycelia from different Se concentration and control groups.
Change
|
Number of metabolites
|
0.1 mmol/L vs. CK
|
0.2 mmol/L vs. CK
|
Increase
|
3
|
2
|
Decrease
|
35
|
29
|
Table 4
Changes in metabolite profiles in Epichloë sp. mycelia between Se and control groups at weeks 4∼8.
Change in
Se vs. CK
|
Number of metabolites
|
week 4
|
week 5
|
week 6
|
week 7
|
week 8
|
Increase
|
42
|
21
|
22
|
24
|
30
|
Decrease
|
86
|
98
|
90
|
109
|
91
|
Marked metabolites changed in Epichloë sp. fermentation broth under selenium condition
To identify metabolite features that were significantly different between each treatment and the control, One Way ANOVA statistical analysis (P < 0.001) were performed. The heatmap visualization showed different trends of metabolite changes for each time of exposure to Se concentrations and different time of exposure to each Se concentration (Figs. 8 and 9). Comparison of the metabolites of fermentation broth exhibited significant differences among culture time and Se concentrations. 64, 40, 36, 49 and 30 distinct metabolites in fermentation broth were identified respectively in weeks 4, 5, 6, 7 and 8 (Fig. 8). Thirteen common metabolites among culture time were pyruvic acid, 2-ketoisovaleric acid, sulfuric acid, O-methylthreonine, glycine, succinic acid, threitol, 3-hexenedioic acid, α-ketoglutaric acid, allose, ribose, ornithine, and tagatose.
At Se values of 0, 0.1, and 0.2 mmol/L, respectively, 67, 47 and 39 marked metabolites in fermentation broth were detected using statistical analysis (P < 0.001) during culture process (Fig. 9). There were 25 common metabolites among Se concentrations, such as pyruvic acid, lactic acid, glycolicaccid, 2-ketoisovaleric acid, sulfuric acid, lactamide, ethanolamine, phosphate, glycine, 2-deoxytetronic acid, L-malic acid, threitol, oxoproline, 4-aminobutyric acid, tartaric acid, allose, ribose, xylitol, D-arabitol, putrescine, 3,6-anhydro-D-galactose, glucose-1-phosphate, O-phosphorylethanolamine, lactose, and maltose.
0.1 or 0.2 mmol/L Se addition in the medium significantly increased some amino acids of fermentation broth (P < 0.001), such as alanine and citrulline in week 4, tyrosine in week 5, serine in both weeks 4 and 5, oxoproline and phenyalanine in the 4th and 6 to 7 week, isoleucine in the 4th, 7th and 8th weeks, aspartic acid in 4 to 7 week, valine in the 4th, 5th, 7th and 8th weeks, glycine and ornithine over the course of the experiment, compared to those of the control fermentation broth. Besides, there were substantial time changes in several amino acid concentrations of fermentation broth (Fig. 10). Highly significant (P < 0.001) effects of culture time were detected for glutamine and O-methylthreonine of the control fermentation broth, proline, tyrosine, and serine of 0.1 mmol/L Se treated fermentation broth, ornithine, phenylalanine, and isoleucine levels of fermentation broth under control and 0.1 or 0.2 mmol/L Se, and glycine, aspartic acid, oxoproline, 4-aminobutyric acid levels of fermentation broth with all experimental media (0ཞ0.2 mmol/L Se). Regarding given treatment time point variance, those levels of both fermentation broth added with 0.1 or 0.2 mmol/L Se were higher in the 4th week than in other time points, however those levels of control fermentation broth were higher in weeks 6 or 8 than in other time points.
During the incubation with all tested selenium concentrations (0.1ཞ0.2 mmol/L Se), the pyruvic acid, 2-ketoisovaleric acid, and α-ketoglutaric acid of the fermentation broth were much more than those in the control medium (P < 0.001), whereas succinic acid and 3-hexenedioic acid in fermentation broth were lower than those in the non-treated ones (P < 0.001). Furthermore, there was time effect on those compound contents (Fig. 10). In case of all culture medium, significant increases in 2-ketoisovaleric acid and 3-hexenedioic acid were observed after 8 week cultivation. Pyruvic acid for the control group was much higher in the 4th week than after 5 week cultivation, however pyruvic acid in 0.1 mmol/L Se culture broth was greater in the 7th week than in the 8th week and before treatment (P < 0.001), and pyruvic acid level of fermentation broth with 0.2 mmol/L Se was sharply increased in the last week of cultivation (P < 0.001). A significant increase in succinic acid was observed in the 5th week in experimental media supplemented with 0 and 0.1 mmol/L Se culture media (P < 0.001). Control fermentation broth and fermentation broth with 0.1 mmol/L Se for α-ketoglutaric acid levels were significantly (P < 0.001) increased in the 4th week compared to other time points, and α-ketoglutaric acid levels in 0.2 mmol/L Se fermentation broth markedly enhanced in the 5th week (P < 0.001). The lactic acid concentrations of control fermentation broth were much higher in the 5th week than other time points, lactic acid of 0.1 mmol/L Se treated fermentation broth were greater in weeks 7 and 8 of cultivation than before 6 week cultivation (P < 0.001), and lactic acid concentrations for 0.2 mmol/L Se treated fermentation broth peaked in the 7th week. Glycolicaccid of control fermentation broth were significantly increased within 5ཞ6 weeks of cultivation compared to other time points (P < 0.001), the glycolicaccid levels in fermentation broth with 0.1 mmol/L Se sharply increased in week 4 (P < 0.001), and glycolicaccid levels of 0.2 mmol/L Se treated fermentation broth were much more in 4 and 6 weeks of cultivation than other time points (P < 0.001). Highly significant increases in 2-deoxytetronic acid were respectively observed in the 6th week in control experimental media and in the 4th week in experimental media supplemented with 0.1 and 0.2 mmol/L Se culture media (P < 0.001). Fermentation broth without addition of Se had significantly higher L-malic acid in week 5 than other culture time (P < 0.001), fermentation broth for 0.1 mmol/L Se had much more L-malic acid in 7 to 8 weeks than before treatment (P < 0.001), and the maximum L-malic acid of fermentation broth in the presence of 0.2 mmol/L Se were observed during the last week of cultivation (P < 0.001). The tartaric acid levels for the control and 0.2 mmol/l Se-treated fermentation broth were greater in week 8 than before cultivation time (P < 0.001), however the tartaric acid levels in 0.1 mmol/L Se-treated fermentation broth were much lower in the 6th week than other time points (P < 0.001).
Fermentation broth with Se significantly accumulated tagatose (P<0.001) compared to the control fermentation broth, however greatly decreased allose, ribose, and threitol. Many sugars and sugar alcohols of fermentation broth were significantly different across the time points (P < 0.001). Fermentation broth for the control and 0.2 mmol/L Se showed higher allose values in the 4th week than after 5 week cultivation, and 0.1 mmol/L Se treated fermentation broth had greater allose in week 5 than other time points (P < 0.001). Maltose of fermentation broth without selenium peaked in weeks 4 and 5 of cultivation. The maltose levels of fermentation broth with addition of 0.1 mmol/L Se was significantly higher in the 4th and 7th weeks than other time points, and maltose levels of fermentation broth with addition of 0.2 mmol/L Se was significantly less in the 7th week than before treatment or in the 8th week (P < 0.001). Lactose levels of control fermentation broth were significantly higher in week 7 than before treatment and in the last week, fermentation broth in the presence of 0.1 or 0.2 mmol/L Na2SeO3 were greater in week 5 than other time points (P < 0.05). The ribose concentrations for control, 0.1 mmol/L Se, and 0.2 mmol/L Se treated fermentation broth peaked respectively in the 4th, 5th and 6th weeks (P < 0.001). The highest xylitol of the control fermentation broth was observed in the 6th and 8th weeks, the lowest xylitol in fermentation broth with the addition of 0.1 or 0.2 mmol/L selenium were obtained in weeks 6 and 7 of cultivation, respectively. The increase in D-arabitol in all fermentation broth were found to be the highest during 8 week of cultivation.
Marked metabolites changed in Epichloë sp. mycelia under selenium condition
As shown in Figs. 10 and 11, there were significant differences in metabolites of mycelia between culture time and Se concentration. In weeks 4, 5, 6, 7, and 8, respectively, 17, 27, 16, 41 and 6 distinct metabolites were identified in mycelia using LSD (P < 0.05). At Se values of 0, 0.1, and 0.2 mmol/L, respectively, 6, 18 and 44 marked metabolites were found in mycelia at a significance level P < 0.05. There were not common marked metabolites across the time points (P < 0.05) or different Se concentrations. Over the time course, Se promoted N-acetyl-β-D-mannosamine and glucoheptonic acid levels, but inhibited 19 metabolites such as glycolic acid, oxalic acid, 3-hydroxybutyric acid, sulfuric acid, malonic acid, hydroxyurea, dihydroxyacetone, ethanolamine, phosphate, proline, glycine, succinic acid, uracil, 2-deoxytetronic acid, 4-aminobuyric acid, 4-hydroxyphenylethanol, conduritol b epoxide, gentiobiose, and isomaltose.
There were no difference in serine, homoserine, methionine, oxoproline, glutamic acid, phenylalanine, ornithine, N-methyl-alanine, cycloleucine under different Se treatments at a specific culture time. Se at 0.1 and 0.2 mmol/L concentrations significantly increased only citrulline level in the 4th week, but significantly decreased tyrosine level in the 5th and 7th weeks and N-ethylglycine in the 5th week. In the 7th week, alanine, valine, isoleucine, proline, asparagine levels of mycelia with 0.2 mmol/L Se were significantly lower than those of other mycelia, and glycine, serine, threonine and carnitine levels for 0.2 mmol/L Se-treated mycelia were also lower than those of the control mycelia. Similarly, aspartic acid level of mycelia with 0.1 mmol/L Se were lower than those of the control mycelia in week 5. Significant time effects on alanine, valine, isoleucine, proline, glycine, serine, threonine, asparagine, aspartic acid, methionine, oxoproline, glutamic acid, phenylalanine, citrulline levels of mycelia with 0.2 mmol/L Se or ornithine levels of mycelia for 0.1 mmol/L Se were observed (P < 0.05, P < 0.001). Those amino acid levels of mycelia were increased in week 5 of cultivation compared to other time points.
Se concentration was the significant (P < 0.05) effect detected on some organic acids for endophytic mycelia (Fig. 11). In case of both 0.1 and 0.2 mmol/L Se culture medium, increases in 2-ketoisovaleric acid, tartaric acid, and lignoceric acid were observed in the 4th, 7th, and 8th weeks, respectively. Decreases in benzoic acid, galactonic acid, and 3-hydroxybutyric acid were found in mycelia treated with 0.1 and 0.2 mmol/L after 5 week or 7 week cultivation. In the 5th week, α-ketoisocaproic acid, cis-gondoic acid, and palmitic acid levels of mycelia with 0.2 mmol/L Se were significantly greater than those of other mycelia (P < 0.05), but arachidonic acid levels of mycelia with 0.1 mmol/L Se were less than those of other mycelia. In week 6, 0.1 mmol/L Se significantly inhibited malonic acid and pipecolinic acid levels in mycelia (P < 0.05). Similarly in the 7th week, 3-methyglutaric acid, 2-deoxytetronic acid, and oxalacetic acid levels of mycelia added by 0.2 mmol/L Se were much greater than those of other mycelia (P < 0.05). During the last week, 0.2 mmol/L Se enhanced xanthurenic acid levels of mycelia. With respect to the time series, oxalacetic acid levels of the control mycelia were lower in the 4th week than after 5 week cultivation (P < 0.05). In addition, mycelia treated with 0.1 mmol/L Se had the lowest 1-monopalmitin in week 4, demonstrated higher lignoceric acid in the 7th week than in the 8th week and before treatment. The lactic acid, tartronic acid, α-ketoisocaproic acid levels of mycelia with addition of 0.2 mmol/L Se were higher in week 4 of cultivation than after treatment (P < 0.05). The lowest succinic acid and pipecolinic acid in mycelia with the addition of 0.2 mmol/L selenium were obtained in the last week. The 0.2 mmol/L Se treated mycelia showed lower 2-deoxytetronic acid in weeks 7 and 8 of cultivation than in weeks 4 and 5 of cultivation.
The presence of selenium decreased significantly gentiobiose, leucrose, and turanose in the 4th or 5th weeks. The concentrations of maltose, isomaltose, lactose, and trehalose at dose of 0.1 mmol/L Se were equal to or above the control mycelia at certain time points, and galactose and ribose were significantly lower for 0.1 mmol/L Se-treated mycelia than those of the control mycelia in the 6th and 8th weeks, respectively. When analyzing time shifts in sugars, mycelia treated with 0.1 mmol/L Se had significantly lower sucrose in the 5th week than other time points, and lower maltotriose in the 8th week than before treatment (P < 0.05), however higher maltose and trehalose in week 4 than after 5 week cultivation. In addition, the greatest lactose and cellobiose in mycelia added by 0.2 mmol/L Se were obtained in week 4 of cultivation.
There were significant (P < 0.05) effects of Se concentration on sugar alcohols in certain time points. Sorbitol of mycelia was higher under 0.2 mmol/L Se than under control and 0.1 mmol/L Se in week 4. Galactinol and lanosterol derived from mycelia were lower under 0.2 mmol/L Se than under control and 0.1 mmol/L Se in the 7th week. Ergosterol concentration of mycelia with the addition of 0.1 mmol/L Se were significantly higher than those of other mycelia in the 7th week. Additionally, there was significant time effect on some sugar alcohols (P < 0.05). Mycelia in the presence of 0.2 mmol/L Se had lower galactinol in weeks 7 and 8 of cultivation than in weeks 4 and 5 of cultivation, lower sorbitol in week 7 than other time points, and higher glycerol in the 5th and 6th weeks than other time points.
Significant effect of Se treatment was detected for some bases and nucleosides for endophytic mycelia. There were marked inosine increases in control mycelia in the 4th week (P < 0.05). A small decrease in thymidine of mycelia was observed in the 5 week cultivation in experimental media supplemented with 0.1 mmol/L Se culture media. Selenium at 0.2 mmol/L concentrations culture media significantly reduced thymine and guanosine of mycelia in the 7th or 6th weeks. Moreover, there were substantial time changes in bases and nucleosides. Guanine levels of 0.1 mmol/L Se treated mycelia were less in the 4th week than after 5 week cultivation. Uracil and thymidine levels for 0.2 mmol/L Se treated mycelia peaked in the 5th week. The thymine levels of mycelia treated with 0.2 mmol/L Se was significantly less in the 7th and 8th weeks than in the 4th and 5th weeks (P < 0.05). In case of both 0.1 and 0.2 mmol/L Se culture medium, increases in guanosine were observed in the 4 week cultivation.