PaGLK transgenic lines showed different gene expression patterns
To explore the transcriptional expression characteristics of PaGLK transgenic poplar, the leaves of PaGLK overexpressing lines, PaGLK suppressing lines, and WT were subjected to RNA-seq sequencing. Due to abnormal clustering of samples OE3 and WT3, these outliers were removed, and a total of 307.9 million Clean Reads were obtained (39.8 ~ 47.9 million per library), with a Q30 base percentage above 92.97% and alignment rate to the Populus trichocarpa genome between 74.10% and 76.63% (Table S1), indicating reliable transcriptome data.
Principal Component Analysis (PCA) showed that the clustering position of WT was closer to that of the overexpressing lines and far from that of the suppressing lines (Fig. 1a). OE vs WT had 3052 DEGs, 1731 upregulated and 1321 downregulated. in RE vs WT had 1326 DEGs, 669 upregulated and 657 downregulated. OE vs RE had 4577 DEGs, 2376 upregulated and 2201 downregulated (Fig. 1b). There were more upregulated DEGs than downregulated genes in all three comparison groups, with RE vs WT having the least DEGs, and the up/downregulated gene ratio in OE vs WT being uneven.
There were 597 common DEGs (317 upregulated, 280 downregulated) in RE vs WT and RE vs OE. There were also 1705 common DEGs (942 upregulated, 763 downregulated) in OE vs WT and RE vs OE (Fig. 1c).
Functional analysis of DEGs
GO functional enrichment analysis was performed mainly for biological processes on the DEGs in the three comparison groups. The common upregulated DEGs in RE vs WT and RE vs OE were enriched related to salt stress response, response to darkness, carbohydrate response, and water shortage (Fig. 2a). The downregulated DEGs were related to photosynthetic electron transfer in photosystem I, photosynthesis, light harvesting in photosystem I, response to light stimulus, chloroplast development, and chloroplast organization (Fig. 2b).
The upregulated DEGs in OE vs WT and OE vs RE were related to cytokinesis during male meiosis, spindle assembly checkpoint signaling in mitosis, DNA replication initiation, and protection of germ cell differentiation (Fig. 2c). The downregulated DEGs were related to regulation of cell fate determination, response to bacterial derived molecular patterns, protein autophosphorylation, and salicylic acid activation signaling pathway (Fig. 2d). These results reveal that the photosynthetic capacity of the RE transgenic lines was negatively affected compared to the WT.
Pathway analysis showed that the DEGs in OE vs WT and RE vs WT corresponded to 120 and 92 KEGG pathways, respectively. The photosynthesis-antenna proteins, photosynthesis, and porphyrin and chlorophyll metabolism pathways were significantly enriched in RE vs WT (Fig. 3a). The ribosome pathway was significantly enriched in OE vs WT (Fig. 3b).
Validation of DEGs by RT-qPCR
We found that downregulated DEGs in RE vs WT and RE vs OE were enriched in pathways related to photosynthesis, and focused on the DEGs in six biological processes related to photosynthesis. We chose ten genes from these processes for qPCR validation and found that the expression levels of these genes in the WT lines were higher than those in the RE lines (Fig. 4).
Metabolome analysis
The response intensity and retention time of each chromatographic peak in the QC sample total ion flow map overlapped substantially (Figure S1), indicating low variation due to instrumental error. A total of 5128 ion peaks were extracted from the metabolites, 2583 positive ions and 2545 negative ions. All extracted peaks were used for PCA analysis.
PCA analysis of the metabolites showed a certain separation tendency on the PC1 and PC2 axis plots, and the distribution of the sample points was somewhat discrete, indicating differences in metabolites among the three groups of samples (Fig. 5).
Analysis of DEMs
KEGG pathway enrichment analysis of differential expressed metabolites (DEMs) found a total of 39 enriched differential metabolic pathways in RE vs WT. Ten of them were significantly enriched (P < 0.01), and mainly related to amino acid metabolism, carbohydrate metabolism, and fatty acid metabolism (Fig. 6a). A total of 49 enriched differential metabolic pathways were found in OE vs WT, and 15 of them were significantly enriched (P < 0.01), mainly related to amino acid metabolism and carbohydrate metabolism (Fig. 6b).
Cluster analysis was performed on the metabolites with significant differences in the metabolic pathways involved in the three comparison groups (Fig. 7). Compared with the WT, more of the downregulated metabolites were found in the PaGLK overexpression lines, mainly including biphenols (catechol, epinephrine, 3,4-dihydroxyphenylethylene glycol, 3,4-dihydroxyphenylethanoic acid), methoxyphenols (vanillin, 3-methoxytyramine), and carbohydrates and carbohydrate-associated molecules (gomphrenin-I, arbutin). In the PaGLK suppression lines, the levels of dicarboxylic acids and their derivatives (fumaric acid, malic acid, succinic acid), eicosanoic acids (prostaglandin D2, prostaglandin G2, PGB2), and amino acids, peptides and analogues (yeast amino acid, O-carbamyl-D-serine) were upregulated, while the levels of these metabolites were downregulated in the WT. These results indicate that differential metabolites have different expression patterns in the transgenic and WT poplar.
Transcriptome and metabolome conjoint analysis
Pearson correlation analysis was done between transcriptome and metabolome. DEGs and DEMs were selected with threshold R-value ≥ 0.95 and P-value ≤ 0.01 for nine quadrant analysis. There were no correlated genes in either the positive or negative ion metabolites in the first quadrant. In the third quadrant, there were 236 regulatory relationships between positive ion metabolites and genes, and 193 regulatory relationships between negative ion metabolites and genes. In the seventh quadrant, there were 72 regulatory relationships between positive ion metabolites and genes, and 69 regulatory relationships between negative ion metabolites and genes. In the ninth quadrant, there were 19 regulatory relationships between positive ion metabolites and genes, and 15 regulatory relationships between negative ion metabolites and genes (Fig. 8). The results indicate that the DEMs have a positive regulatory relationship with DEGs.