2.1. Qualitative and Quantitative Metabolites
A total ion current (TIC) plot and multimodal detection plot for a quality control (QC) sample are shown in Figure 1a. The TIC plot represents a continuous description of the intensity and continuity of all ions in the mass spectrum at different time points. The curves for the total ion flow overlap greatly, and the retention time and peak intensity are the same, indicating better signal stability for the same sample when detecting mass spectrometry at different times.
The multimodal detection plot of metabolites in multiple reaction monitoring (MRM) mode shows that our sample contains multiple substances. Each mass-spectrometric peak of a different color represents the detected metabolite. Based on the local metabolite database, qualitative and quantitative mass spectrometry analysis of the metabolites in the samples was performed. A total of 752 metabolites were identified, including 155 phenolic acids, 119 flavonoids, 38 sugars and alcohols, 57 lipids, 66 amino acids and derivatives, 51 organic acids and derivatives, 36 nucleotides and derivatives, 33 lignin and coumarin, 2 quinones, 50 alkaloids, 19 terpenes, 15 vitamins and derivatives and 15 other metabolites. Details of all identified metabolites are shown in Table S1.
2.2. PCA and PCC Analysis
In order to compare the metabolite composition of Sik. under four different cold stresses, the principal component analysis (PCA) was performed . PCA was performed on samples, and PCA provides insight into overall metabolic differences between samples and within-group variation (29). The PCA plot shows that the Sik. under five conditions is clearly separated and the three biological replicates of each condition are clustered together (Figure 1b), , indicating that the experiment is reproducible and reliable. The plants were significantly separated from the control CK group in the four low-temperature treatments, and there were also different degrees of separation between the four groups, indicating that there were significant differences in metabolite content between each group, which in turn indicated that the accumulation of different metabolites occurred after being subjected to different conditions of low-temperature stress. The T1/T3 group and the medium PCA map also showed obvious separation (Figure 1c), indicating that the reaction mechanism of low temperature and freezing temperature were different, and the T2/T4 group and the medium PCA map also showed obvious separation indicating that the metabolites accumulated after cold domestication were also different (Figure 1d), and the obvious separation between the T2/T3 group and the medium PCA map indicated that the metabolites also changed under freezing and extreme freezing conditions (Figure 1e).
The Pearson correlation coefficient can be used to observe biological replicates between samples within groups by analysis of the correlation between samples. At the same time, the higher the correlation coefficient of the samples in the group relative to the samples between the groups, the more reliable the different metabolites obtained. Generally speaking, when 0.8≤|r|≤1, it can be considered that the two variables are highly correlated, 0.5≤|r|≤0.8 moderately correlated, 0.3≤|r|≤0.5 low correlation, the experimental data shows that the Pearson correlation coefficient in each group is 0.8≤|r|≤1, and the samples between groups are also at 0.5≤|r| indicating that the biological reproducibility in each group is high and the differential metabolites are reliable (Figure S1). These results showed that the experimental data were reliable and that different low temperature conditions strongly affected the metabolite profile of snow lotus.
2.3. Orthogonal Partial Least Squares-Discriminant Analysis
To find differential metabolites, orthogonal partial least squares discriminant analysis (OPLS-DA) was used to extract components in the independent variable X and the dependent variable Y, and then correlations between components were calculated. OPLS-DA combines orthogonal signal correction (OSC) and PLS-DA methods to decompose the X matrix information into two types of correlated and uncorrelated differences with Y. The differential variables were then screened by removing the uncorrelated differences. The results showed that all R2X were higher than 0.54. All R2Y scores were higher than 0.99 and all Q2 values were greater than 0.84 at CK vs T1/T2/T3/T4, respectively (Figure 2a-d), confirming that the differential metabolites responded to low temperature stress treatment. The S-plot map of OPLS-DA is shown in Figure 2 e-h. The OPLS-DA model was also validated by 200 random alignment and combinatorial alignment tests.
2.4. Differential Metabolite Screening
Based on fold change≥2 or ≤0.5 and VIP≥1, a total of 362 differential metabolites (DEMs) were detected in the CK vs T1/T2/T3/T4 group (Table S2). Venn plots show common and unique differences between cold treatment and control metabolomes(Figure 3a-c). The results showed that in the overlapping regions of the Venn diagram, 72 upregulated differential genes and 6 downregulated differential genes were obtained, respectively. Volcano maps of differential metabolites in different pairwise comparisons are shown in Figure 3d-j. In the CK vs T4 group, there were more metabolites involved in cold regulation. There were 221 significantly different metabolites (163 up-regulated, 58 down-regulated) between CK vs T1, 171 significantly different metabolites (146 up-regulated, 25 down-regulated) between CKvsT2, 186 markedly different metabolites (135 up-regulated, 51 down-regulated), and 240 significantly different metabolites (173 up-regulated, 67 down-regulated) between CK vs T4. Interestingly, all upregulated differential metabolites were detected in the T2 cryogroup overlapping with the other groups, and only 6 upregulated differential metabolites were unique. There were 127 significantly different metabolites (53 up-regulated, 74 down-regulated) between T1vsT3 groups, 165 markedly different metabolites (67 up-regulated, 98 down-regulated) between T2 vs T4, and 115 significantly different metabolites (36 up-regulated, 79 down-regulated) between T2 vs T3. These metabolites are divided into 10 groups: amino acids and their derivatives, phenolic acids, lipids, organic acids, nucleotides and their derivatives, lignans and coumarins, flavonoids, alkaloids, terpenes and others. Metabolite changes were concentrated in Lipids, Phenolic acids, Flavonoids, Nucleotides and derivatives, Amino acids and derivatives (Table 1). In particular, more than half of these differential metabolites are secondary metabolites, including phenolic acids, flavonoids, flavonols, dihydroflavones, isoflavones, coumarins, and alkaloids.
2.5. Functional Annotation and KEGG Enrichment
The annotated results of the significantly different metabolites KEGG were classified by pathway type in KEGG, and the differential metabolites in the CK vs T1/T2/T3/T4 group involved metabolic pathways, secondary metabolite biosynthesis, cofactor biosynthesis, amino acid biosynthesis, and purine metabolism. In addition to the above, T1vsT3 is also involved in the biosynthesis of flavonoids, T2 vs T4 is involved in nucleotide metabolism, linolenic acid metabolism, and T2 vs T3 is involved in flavonoid metabolism. Interestingly, the proportion of metabolites in the ATP binding cassette (ABC) transport pathway increased significantly in all three comparison groups (Figure S2). Based on the results of differential metabolites, the KEGG pathway of DEMs was performed on four different conditions of Sik.. The main metabolic pathway changed in the CK vs T1 group are starch and sucrose metabolism, C5-branched dibasic acid metabolism, galactose metabolism, etc.; when the temperature drops to -13℃, the more obvious channels are purine metabolism, starch and sucrose metabolism, and biosynthesis of cofactors; CK vs T3 from which it can be seen that compared with the CK vs T1 group, the pathway change is significantly reduced, but in the comparison of sub-zero low temperature, we find isoflavonoid biosynthesis, purine metabolism, glycerophospholipid metabolism, photosynthesis, etc. After cold acclimation, and the significantly changed pathways are Starch and sucrose metabolism, vitamin B6 metabolism, glycerophospholipid metabolism, C5-Branched dibasic acid metabolism, and metabolism of glycine, serine and threonine. The differential metabolic pathways under cold and freezing temperature (T1 vs T3) conditions were mainly focused on indoleic acid biosynthesis, isoflavone and flavonoid biosynthesis pathways. The differences between the freezing and cold-tamed groups (T2 vs T4) mainly showed significant differences in linoleic acid metabolism, isoflavone biosynthesis and nucleotide metabolic pathways, and the differential metabolic pathways in the different freezing temperature groups (T2 vs T3) were mainly purine metabolism, isoflavone biosynthesis and nucleotide metabolic pathways (Figure 4a-g).
Table 1. The amount of differential metabolites in Sik. leaves under low temperature stress.
Group Class
|
CK/T1
|
CK/T2
|
CK/T3
|
CK/T4
|
T1/T3
|
T2/T4
|
T2/T3
|
Up
|
Down
|
Up
|
Down
|
Up
|
Down
|
Up
|
Down
|
Up
|
Down
|
Up
|
Down
|
Up
|
Down
|
Amino acids and derivatives
|
7
|
5
|
5
|
5
|
5
|
4
|
15
|
5
|
2
|
4
|
13
|
0
|
4
|
2
|
Phenolic acids
|
32
|
9
|
25
|
5
|
13
|
9
|
30
|
12
|
8
|
22
|
9
|
6
|
8
|
13
|
Nucleotides and derivatives
|
12
|
3
|
12
|
0
|
11
|
3
|
9
|
2
|
3
|
2
|
10
|
1
|
2
|
7
|
Flavonoids
|
31
|
16
|
11
|
8
|
22
|
11
|
33
|
14
|
8
|
15
|
13
|
11
|
9
|
8
|
Lignans and Coumarins
|
6
|
2
|
7
|
1
|
5
|
3
|
6
|
1
|
1
|
5
|
2
|
1
|
2
|
3
|
Alkaloids
|
4
|
11
|
5
|
3
|
5
|
4
|
3
|
9
|
2
|
3
|
3
|
6
|
3
|
5
|
Terpenoids
|
4
|
0
|
4
|
0
|
4
|
0
|
4
|
2
|
0
|
1
|
0
|
1
|
1
|
2
|
Organic acids
|
8
|
6
|
4
|
0
|
4
|
3
|
8
|
4
|
2
|
3
|
4
|
2
|
1
|
5
|
Lipids
|
37
|
5
|
58
|
2
|
54
|
11
|
47
|
10
|
26
|
6
|
9
|
63
|
6
|
26
|
Others
|
22
|
1
|
15
|
1
|
12
|
3
|
18
|
8
|
1
|
13
|
4
|
7
|
0
|
8
|
Total
|
163
|
58
|
146
|
25
|
135
|
51
|
173
|
67
|
53
|
74
|
67
|
98
|
36
|
79
|
2.6. Differential Metabolites Analysis
2.6.1. Low-Temperature-Induced Lipid Changes
A total of 133 lipids, 59 free fatty acids, 25 lysophosphatidylethanolamine (LPE) and 27 lysophosphatidylcholine (LPC) were detected at low temperatures. Compared with the control group, the expression of multiple substances in LPE and LPC was increased by more than 10 times in the T1/2/3/4 group (Figure S3), and the detected LPE and LPC were increased or did not change significantly, except for the decrease in the content of LysoPC 20:4 in CK vs T3/4 group. The only detected glycophosphate choline (GPC) was upregulated in all four groups. However, the lipid content of T2vsT4 (Table S5) and T2vsT3 (Table S4) decreased significantly. No significant lipid changes in T1 vs T3 group (Table S3).
2.6.2. Low-Temperature-Induced Changes in Phenylpropanoid Pathway
Phenolic acids are a secondary metabolite containing phenolic rings. In this study, we identified 11 phenolic acid compounds with significant differences in all four groups compared with CK under four low temperature conditions in Sik. leaves, including 4-nitrophenol, caffeic acid, salicylacetic acid, sinapinaldehyde, chlorogenic acid methyl ester, caffeoylferuloylquinic acid, 3,4-O-dicaffeoylquinic acid methyl ester, 1,4-O-dicaffeoyl-3-O-succinoyl-quinic acid, 1,5-O-dicaffeoyl-3-O-succinoyl-quinic acid, dicaffeoylquinic acid-O-glucoside, 3,5-di-O-caffeoyl-1-O-(2,4-disuccinoyl)-quinic acid, except for 4-nitrophenol and dicaffeoylquinic acid-O-glucoside, are all upward. Most of these metabolites are involved in the phenolic acid branch of phenylpropanoid metabolism and are further involved in flavonoid biosynthesis. Three coumarins 7-hydroxycoumarin-O-rhamnoside, 4-hydroxy-7-methoxycoumarin-β-rhamnoside, scopoletin-7-O-glucuronide, which showed significant changes; 10 flavonoids, namely pinocembrin, 4',5-Dihydroxy-3', 6,7-trimethoxyflavone, tricetin, isohyperoside, quercetin-3-O-galactoside, quercetin-7-O-glucoside, quercetin-3-O-glucuronide, Isorhamnetin-3,7-O-diglucoside, formononetin, ononin, sesquiterpenoids, reynosin, and 5α-hydroxycostic acid all showed an upward trend. Four substances were adjusted downward, among which pinocembrin, 4',5-dihydroxy-3',6,7-trimethoxyflavone belonged to dihydroflavones, and formononetin, ononin belonged to isoflavones, while the rest of flavonoids and flavonols were adjusted upward. The massive accumulation of these flavonoids in snowdrop leaves may act as endogenous antioxidants in the plant defense mechanism under low temperature stress. Twenty metabolites of this pathway were identified, and in addition to some of the above metabolites there wereidentified, including quinic acid, protocatechuic acid, shikimic acid, phenylalanine, 4-hydroxybenzoic acid, salicylic acid, cinnamic acid, coniferyl alcohol, pinoresinol, prunetin, esculetin, coumarin, apigenin, salicylacetic acid, protocatechuic acid, dihydrochrysin. These metabolites were upregulated except for Prunetin, which was downregulated in four groups, Phenylalanine, which was downregulated in CK vs T1 and CK vs T2, and Pinoresinol, which was downregulated in CK vs T2, CK vs T3, and CK vs T4. In the comparison group of T1 vs T3, T2 vs T3 and T2 vs T4, these substances were in an upward trend as a whole. Phenylalanine and sinapinaldehyde declined in all three comparison groups. It indicates that more kinds of phenolics can be produced under freezing conditions at -13 °C than at -4 °C, and more phenolics can be produced after cold domestication.
2.6.3. Low-Temperature-Induced Changes in Purine Metabolism Pathway
Nucleotides are nucleic acid hydrolysates, and in this study, nucleotides and their derivatives were identified in the four comparison groups: guanine 5-aminoimidazole ribonucleotide, 2'-deoxyinosine-5'-monophosphate, guanosine 3',5'-cyclic monophosphate, and uridine 5'-diphosphate both are up-regulated. These metabolites are all involved in the purine degradation pathway. We identified 9 nucleotides and their derivatives with significant low-temperature-induced accumulation in Sik. leaves, including Guanosine, Guanine, Hypoxanthine, Xanthine, Adenosine, Guanosine 3',5'-cyclic monophosphate, Adenosine 5'-diphosphate, Cyclic 3',5'-Adenylic acid、2'-Deoxyadenosine.
Guanosine, guanine, and hypoxanthine showed a significant up-regulation trend under low-temperature stress. Compared with the CK group guanine significantly increased in all four groups of low temperature stress, while Guanosine, Hypoxanthine showed a significant increase in the T2 group. Xanthine, adenosine showed an overall down-regulation trend, adenosine significantly decreased in the T3 group. Interestingly, xanthine in T3 group was upregulated and adenosine was upregulated in the T4 group. The intermediate adenosomes guanosine 3',5'-cyclic monophosphate, adenosine 5'-diphosphate showed significant accumulation in all four comparison groups, cyclic 3',5'-adenylic acid, 2'-deoxyadenosine showed an overall tendency to decrease. Although the accumulation of intermediate glands was obvious but the end products xanthine and adenosine were significantly reduced, only guanine showed a significant accumulation. The identified nucleotides and their derivatives were upregulated in all three comparison groups except adenosine 5'-diphosphate which was downregulated in T1 vs T3, T2 vs T3, T2 vs T4 groups, guanine and xanthine which were downregulated in T1 vs T3 and T2 vs T3 and hypoxanthine which was downregulated in T1 vs T3. Significant accumulation was observed in the T2 vs T4 group except for adenosine 5'-diphosphate.
2.6.4. Low-Temperature-Induced Glycolysis and TCA Metabolic Pathway
We identified six metabolites associated with glycolysis and TCA metabolic pathways, including D-glucose, D-glucose-6-phosphate, D-fructose-6-phosphate, succinic acid, oxaloacetate, and isocitric acid. Compared with the CK group, except for oxaloacetate in T2 and T3, the other metabolites were on the rise. In the pentose phosphate pathway, ribulose, ribulose-5-phosphate and erythritrose-4-phosphate were significantly accumulated in the four groups. In the synthesis pathway of sucrose and trehalose, the intermediate products D-lactose-6-phosphate and trehalose-6-phosphate were significantly accumulated, and trehalose was in an upward regulation trend except for the T3 group, and there was no significant change in sucrose. In the T1vsT3 group all nine substances were upregulated but in the T2 vs T4 group all nine substances were downregulated. In the T2 vs T3 group, only trehalose, trehalose-6-phosphate, fructose and glucose contents were increased, while the rest of substances were down-regulated or changed insignificantly. It indicates that soluble sugars can be accumulated more after low temperature conditions and cold domestication.
2.6.5. Low-Temperature-Induced Changes in Amino Acids and Their Derivatives
Under low temperature stress, only one amino acid analog, N-acetyl-L-aspartic Acid, was detected to be significantly up-regulated in all four groups compared to the control group.In addition, 12 amino acids and their derivatives were identified. L-cysteine, L-tryptophan and L-asparagine were in a downward adjustment trend in the four groups, while L-tryosine, L-phenylalanine, L-isoleucine and L-leucine decreased in the first two groups and showed an upward trend in the latter two groups. L-serine was down-regulated in the T2 group and up-regulated in the rest of the group. L-methionine was upregulated except for T1, L-ornothine was significantly downregulated in T4, and the rest were upregulated. The remaining amino acids L-lysine, L-proline, and L-grotamine were all upregulated in the four groups. The amino acids detected in the T1 vs T3, T2 vs T3, and T2 vs T4 groups were in a downward trend as a whole, only ornithine increased in the three comparison groups, asparagine, cysteine accumulate in group T2 vs T4, and proline was upregulated in T1 vs T3, indicating that more amino acids could be accumulated after cold acclimation.
2.6. Comprehensive metabolic network analysis under low temperature stress
In order to get a comprehensive understanding of the changes in metabolites under different low temperature stress treatments, we proposed a metabolic pathway based on the literature on metabolic pathways and web databases. The main known pathways include sucrose and alginose synthesis, glycolysis, pentose phosphate pathway, tricarboxylic acid (TCA) cycle, amino acid metabolism, purine metabolism, and phenol metabolism (Figure 5a-b).