1. Agrobacterium adsorbs into holes on the surface of immature wheat embryos.
Ishida et al. developed the high transformation efficiency methods using the immature wheat embryos. In this study, wheat embryos from the co-culture stage of the protocol were selected as the research objects to reveal the effects of Agrobacterium infection. Firstly, we observed by scanning electron microscopy that the upper epidermis of immature wheat embryos would break and create gaps after the pretreatment (Fig. 1a, Fig. 1d). After that, Agrobacterium can accumulate in large quantities between the ruptured cell gaps (Fig. 1b, Fig. 1e). Interestingly, we found that Agrobacterium accumulates little in the smooth cortex but is abundant below the torn cortex (Fig. 1c, Fig. 1f). We believe that Agrobacterium may contact wheat embryonic cells through cracks caused by the pretreatment, thus affecting the transformation efficiency.
Subsequently, some physiological indicators related to oxidative stress were measured (Fig. 2). The results showed that the expression of reductases, including SOD and POD, was up-regulated in the immature wheat embryos treated with Agrobacterium tumefaciens, thereby reducing the ROS response, which would be conducive to the T-DNA transfer of Agrobacterium tumefaciens and the regeneration process of immature embryos. The changes of stress resistance indexes included the decrease of GSH, the increase of proline, and the increase of MDA, which indicated that some stress response occurred in immature wheat embryos.
2. Transcriptomic Analysis of the wheat embryos infected byAgrobacterium
To investigate the molecular basis regarding Agrobacterium infects immature wheat embryos to improve the transformation efficiency, whole-genome transcriptome analysis is employed to analyze the expression of transcriptome genes. Three biological replicates of cDNA libraries were prepared for a total of four samples of control and treatment after Agrobacterium-infected immature embryos. At first, the four cDNA libraries from control and treatment after Agrobacterium-infected immature embryos were prepared in Illumina sequencing from three biological replicates per sample, which generated 73174653、75342269、71580265、72900690 sequences with 10.98、11.3、10.74、10.93G from CK1, CK2, T1, T2 respectively(Table S1).
After the removal of low-quality raw reads, which included empty, too short, too many Ns, we obtained an average of 69898136 ,71543825,65160259,68953701 high-quality sequences for CK1, CK2, T1, T2, respectively. These reads were assembled into 120744 total genes and 146597 total transcripts with an average length of 1797 bp.
A principal component analysis (PCA) suggested close clustering of replicates within the samples and hence obtaining excellent sample data reproducibility (Figure S2). A large quantity of unique sequences from wheat immature embryo should cover a vast majority of genes in this species.
A total of 1570 DEGs were identified in the comparison of CK1 vs. T1(Fig. 3a, 3b, Table S2, S3), GO (Gene Ontology) annotation of these genes can be divided into molecular function (2019 sequences), biological process (1005 sequences) and cellular components (2072 sequences). A total of 1155 DEGs were identified in the comparison of CK2 vs. T2(Fig. 3b, Table S2, S4), and the GO annotations of these genes could be classified into molecular functions (1448 sequences), biological processes (512 sequences) and cellular components (2057 sequences). GO term annotation provided a broad overview of the functional groups of genes cataloged in our wheat embryo transcriptome. In addition, the Venn diagram showed 184 common up-regulated and 271 down-regulated differentially expressed genes in CK1 vs. T1 vs. CK2 vs. T2(Figure S3).
As is shown in Fig. 3c and 3d, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that the 124 DEGs in CK1 vs. T1 of wheat were mainly related to Phenylpropanoid biosynthesis (162, 6.99%), Starch and sucrose metabolism(130, 5.61%), Plant-pathogen interaction(108, 4.66%), Plant hormone signal transduction(100, 4.31%), MAPK signaling pathway - plant(82, 3.54%), Galactose metabolism(60, 2.59%), Flavonoid biosynthesis(58, 2.50%), Amino sugar and nucleotide sugar metabolism(58, 2.50%), Purine metabolism(49, 2.11%), Pyrimidine metabolism(48, 2.07%), while the 120 DEGs in CK2 vs. T2 were enriched in Phenylpropanoid biosynthesis(111, 6.73%), Plant-pathogen interaction(104, 6.30%), Plant hormone signal transduction(94, 5.70%), Starch and sucrose metabolism(83, 5.03%), MAPK signaling pathway - plant(72, 4.36%), DNA replication(42, 2.55%), Protein processing in endoplasmic reticulum(38, 2.30%), Nucleocytoplasmic transport(36, 2.18%), Amino sugar and nucleotide sugar metabolism(35, 2.12%), Galactose metabolism(34, 2.06%).
3. Validation of RNA-Seq Data by qRT-PCR
To determine the accuracy of RNA sequencing results, 9 DEGs from wheat embryos treated with Agrobacterium tumefaciens were selected for qRT-PCR (Figure S4, Table S5). The expression of these genes by qRT-PCR was consistent with the transcriptome sequencing data, which validates the useability of transcriptomic data.
4. Metabolic Analysis of the wheat embryos infected by Agrobacterium
To fully understand the metabolic changes that occured in response to Agrobacterium stimulation of wheat embyro, a nontarget metabolic analysis was performed using UPLC-qTOF-MS, 620 were known metabolites among the 7174 metabolites detected (Table S6), most of the compounds were classified into Carboxylic acids and derivatives(88), Organooxygen compounds(64), Benzene and substituted derivatives(38), Fatty Acyls(35), Cinnamic acids and derivatives(31), Phenols(25), Indoles and derivatives(16), Purine nucleosides(13), Imidazopyrimidines(13), Keto acids and derivatives(13), Steroids and steroid derivatives(13), Pyrimidine nucleotides(12), Prenol lipids(11), Organonitrogen compounds(10).
Principal component analysis (PCA, Fig. 4a, 4b) showed that the same treatments were gathered together, indicating good repeatability between samples, while different treatments were separated from each other, indicating that there were different effects on metabolites between treatments.
We screened differentially accumulated metabolites (DAMs) with a screening criterion of FC > 2/q < 0.05 and compared the metabolite numbers between the two control conditions and the two treatment levels to identify DAMs following Agrobacterium infection of immature embryos. The differentially accumulated metabolites were extracted and placed on a pie chart (Fig. 4c). These DAMs mainly included Organic acids and derivatives (34), Organoheterocyclic compounds (29), Organic oxygen compounds (26), Benzenoids (19), Lipids and lipid-like molecules (17), Phenylpropanoids and polyketides (12), Nucleosides, nucleotides, and, analogues (11), Organic nitrogen compounds (4), Alkaloids and derivatives (1). We identified 248 DAMS from CK1 and T1, of which 154 compounds were increased and 94 compounds were decreased (Table S7, Table S8). 304 DAMS were identified from CK2 and T2, of which 175 compounds were increased and 129 compounds were decreased (Table S7, Table S9). In addition, Venn diagrams were used to identify differential metabolites at 24 h and 48 h treatments. 72 up-regulated common differential metabolites were identified in the comparison of CK1 vs. T1 vs. CK2 vs. T2, and 33 down-regulated common differential metabolites were identified in the comparison of CK1 vs. T1 vs. CK2 vs. T2 (Fig. 4d, 4e, Figure S4).
Comparative analysis of wheat immature embryos treated with Agrobacterium tumefaciens showed significant differences in metabolites, and significant enrichment pathways for CK1 and T1 (49 in total) included Purine metabolism, Phosphonate and phosphinate metabolism, Histidine metabolism, Arginine and proline metabolism, Tryptophan metabolism, Phenylpropanoid biosynthesis, Phenylalanine metabolism, Alanine, aspartate and glutamate metabolism, Vitamin B6 metabolism. Pathways significantly enriched in CK2 and T2 (a total of 54 pathways) included Phenylpropanoid biosynthesis, Purine metabolism, Tryptophan metabolism, Pentose phosphate pathway, Phosphonate and phosphinate metabolism, Glutathione metabolism, Phenylalanine metabolism, Tyrosine metabolism, Monoterpenoid biosynthesis, Glycine, serine and threonine metabolism.
5. Association analysis between DEGs and DACs
To better understand the relationship between genes and metabolites after Agrobacterium infects immature embryos, differentially expressed genes and differentially accumulated metabolites were mapped to the KEGG pathway map simultaneously. The results showed that the same pathway of DEGs and DAMs was enriched into glycolysis, TCA cycle, amino acid biosynthesis, phenylpropionin, lignin and other related pathways As shown in the Fig. 5, many genes were changed in Agrobacterium infected wheat immature embryos through glycolysis metabolism, TCA cycle metabolism and their related pathways, the up-regulated genes include PKM (TraesCS3D02G109600), gapN (TraesCS2D02G197300), gpmI (TraesCS4B02G172700), pyk (TraesCS2D02G561700), RAFS (TraesCS3B02G133400), tktA (TraesCS2D02G073900), IDH1 (TraesCS2A02G205900), GLT1 (TraesCS3A02G266300, TraesCS3B02G299800, TraesCS3D02G266400), POP2 (TraesCS2A02G421400), and down-regulated genes include PDHA (TraesCS6B02G342000), PDHB (TraesCS5B02G290600), DLAT (TraesCS4A02G481800), DLD (TraesCS1D02G109600), ACO (TraesCS6B02G342000), LSC1 (TraesCS5B02G290600), LSC2 (TraesCS7D02G011600), SDHA (TraesCS1D02G109600), fumC (TraesCS2A02G336500), MDH1 (TraesCS4A02G481800), and GSS (TraesCS5B02G104500, TraesCS7D02G431500). The up-regulation of PKM, gapN, and gpmI promoted the accumulation of 2PGA, which promoted the accumulation of quinate and 5-o-(1-carboxyvinyl)-3-phosphoshikimate in the downstream shikimate pathway. Pyk was upregulated by 2-fold in immature embryos treated with Agrobacterium tumefaciens, which is a key enzyme that catalyzed the conversion of PEP to pyruvate, and then promoted the biosynthesis of downstream important osmotic substance betaine. Acetyl-CoA is a hub metabolite linking glycolytic pathway and TCA cycle. Several regulated genes were down-regulated, including PDHA, PDHB, DLAT, and DLD, which suggests that Agrobacterium infection activates the glycolytic pathway of immature wheat embryos, and several metabolites enter the shikimic acid pathway through PEP. Six genes encoding five enzymes in the TCA cycle pathway were downregulated in Agrobacterium-infected embryos. GSH is an important regulatory metabolite and affects the metabolic process in cells, GSH encoding enzyme GSS gene was down-regulated, and the accumulation of its downstream metabolites in immature embryos was increased. The biosynthesis of GSH was mediated by succinyl-CoA and 2-oxoglutarate in TCA cycle. Meanwhile, we also detected accumulation of intermediate metabolites, including γ-Glutamylcysteine, Glutamate, GABA, and Succinate semiphosphate. The results indicated that the TCA cycle pathway participated in the response of immature wheat embryos to Agrobacterium, and finally participated in the GSH biosynthesis pathway through intermediate metabolites. In addition, we also detected changes in some amino acid metabolites, including D-glucuronide, D-serine, L-histidine, Isoleucine, D-Aspartate, β-alanine, tyrosine, etc.
Structural genes such as PAL and 4CL were involved in phenylpropanoid metabolism pathway, and many secondary metabolites in this pathway were changed, including the decrease of phenylalanine and cinnamic acid accumulation and the increase of p-coumarate accumulation. We found that many enzymes and metabolites are involved in the process of infecting wheat embryos by Agrobacterium in related synthetic pathways such as phenylpropionin and lignin (Fig. 6). Structural genes such as PAL and 4CL were involved in the phenylpropanin metabolism pathway, where multiple secondary metabolites are altered, including decreased accumulation of phenylalanine and cinnamic acid and increased accumulation of p-coumarate. Following the phenylpropyl pathway into the lignin pathway, several encoding enzyme genes such as CCR, REF1, F5H, bglX got changed, leading to the accumulation of sinapic acid, cinamaldehyde, and coniferaldehyde. In addition, some metabolites linked through pathways such as phenylpropylene and lignin also changed, Up-regulation of ASMT, the down-regulated of gene encoding bx1-5 and up-regulation of bx8/9 genes led to altered accumulation of melatonin. The expression of these genes reduces the accumulation of DIBOA-glucoside, the key metabolites downstream of this pathway. Benzoic acid is the key substance of adipose biosynthesis and its accumulation in wheat embryo is also increasing. In conclusion, wheat immature embryos infected with Agrobacterium tumefaciens participate in phenylpropanoid and lignin metabolic pathways by regulating some key enzymes, and lead to changes in metabolites. Additionally, some metabolites related to oxidative stress response have been accumulated, indicating that immature wheat embryos have some stress responses through phenylpropanoid and lignin pathways after Agrobacterium stimulation.