Phenotypic characterization of S. lycopersicum seedlings under cadmium stress
In all the varieties studied, the morphology and appearance of metal-treated seedlings was altered, particularly in the roots after cadmium ion treatment. We observed that root elongation was inhibited, and root and shoot in dry weight decreased (Fig.1A and B). Compared to the control, the dry weight of shoot and root of seedlings was decreased by cadmium ion stress treatment by 40.42% and 60.39% at day 7, respectively. Moreover, cadmium stress treatment restrained the root lengths of tomato seedlings, which the root length inhibition rate was in Cd2+-treated plants (43.42%) at day 7 compared with control (Fig.1C and D).
Total Chlorophyll content and lipid peroxidation
Malondialdehyde (MDA), as the primary lipid peroxidation by-product, was used for estimating the level of lipid peroxidation. With the prolongation of treatment time, its content also increased (Fig.2A). Its levels in controls (9.01 + 0.21 nmol/g·FW), were significantly lower than those at 10 μM Cd2+ (14.30 + 0.05 nmol/g·FW) at day 3. Furthermore, an evident and visible increase in lipid peroxidation was observed which occurred in roots of seedlings with 10 μM Cd2+ at day 7.
After 7 days of exposure, the inhibitory effects of Cd on total chlorophyll in tomato leaves were shown in Fig.2B. As tomato seedlings grew, the chlorophyll content of control gradually increased. However, the chlorophyll levels of tomato seedlings treated with Cd2+ decreased significantly under oxidative stress. No significant differences (p>0.05) were found between tomato seedlings treated with Cd2+ and control. Compared to control, the chlorophyll content in tomato seedlings treated with Cd2+ had more sharp decline (by 45.95%) on day 7th.
Antioxidant enzyme activities
In the present research, it was found that POD activity was altered with cadmium treatment (Fig.2C). Actually, POD activity in roots with 10 μM Cd2+ treatments for 3 day reached the peak value, and then declined. However, POD activity in control had peaked on the first day. As shown in Fig. 2C, POD activities were distinctly induced by cadmium ion stress at the beginning of the third day.
Likewise, CAT activity of Lycopersicum esculentum underwent a fluctuating trend following cadmium exposure (Fig.2D). When exposed to cadmium for 3 days, the CAT activity reached a maximum value of 189.74 U·g-1 FW. After 7 days of treatment, great decrease of CAT activity occurred as presented in Fig.2D. Generally speaking, CAT activity initially went up, and then decreased. It was obvious that Cd2+ treatment could enhance CAT activity during the entire processing period.
Effect of Cd on lignin content
Seven-day-old tomato seedlings were cultured in Hoagland nutrient solution with 10 μM Cd2+ for 7 days, and the tomato roots were cut with scissors and collected for lignin assay. Phloroglucinol staining indicating lignin deposition in the vascular tissues of tomato seedling roots was visible after Cd treatment for 7 days. As presented in Fig. 3, the visible lignin content in tomato seedling roots incubated in Cd2+ solution significantly higher than control.
Illumina paired-end sequencing and de novo assembly
mRNA from control and cadmium-treated plants were extracted, which were used to construct cDNA libraries, separately. 20.69, 19.36, 25.24, 19.58, 23.65, and 20.49 million clean sequence reads in total were generated from the six libraries of CK1, CK2, CK3, Cd1, Cd2, and Cd3, respectively (Table 1). There were 25,852 unigenes, of the 31,214 unigenes, had been annotated. 11,124 unigenes ranged in lengths from 300-1000 bp, and 14,227 unigenes with lengths of more than 1000 bp (Table 2).
For annotation and validation of the assembled unigenes, sequence similarity searches were carried out in seven public databases, i.e., the GO, NR, NT, COG, KOG, and PFAM databases. The results manifested that, out of 56,044 unigenes, 18,883 (33.69%), 25,848 (46.12%), 25,852 (46.13%), 10,108 (18.04%), 14,553 (25.97%), 21,648 (38.63%), 20,473 (36.53%) and 25,840 (46.11%) unigenes displayed remarkable similarity to known proteins in the GO, NR, NT, COG, KOG, PFAM, SwissProt and KEGG databases, respectively (Fig.4, Table.2). Together, 31,214 (55.70%) unigenes revealed similarity to known proteins in at least one of these seven databases (Table.2).
Identification and functional enrichment analysis of the DEGs
All sequencing reads were realigned to the unigenes in order to confirm the expression levels of 56,044 de novo assembled unigenes. FPKM values for each gene between CK and Cd-treated groups were performed to obtain the DEGs in response to Cd stress. A total of 1,716 DEGs (1,157 up-regulated genes and 559 down-regulated genes) had been identified between CK and Cd-treated groups (Fig. 5).
In our study, GO and KEGG pathway were used to analyze and classify the DEGs. The numbers of DEGs in CK and Cd-treated groups were summarized in three main GO categories (Fig. 6, Table S1). Gene Ontology analyses were performed in seedlings of tomato plants in CK and Cd-treated group. In total, 18,883 genes were matched to 51 GO items, which consisted of 15 items in Cellular Component (CC), 12 items in Molecule Function (MF) and 24 items in Biological Process (BP). Given the negative logarithm of the P value, the major four biological processes identified for these genes were cellular process, metabolic process, biological regulation, and response to stimulus (Fig. 6). The largest four cellular components associated with these genes were cell, cell part, organelle, and membrane (Fig. 6). The major four molecular functions for these genes were binding, catalytic activity, transcription regulator activity, and transporter activity (Fig. 6).
For the COG analysis, 10,108 unigenes (Table 2) from all unigenes were categorized to 20 categories (Fig. 7, Table S2). The most distributed one was the cluster of general function prediction only (163), followed by secondary metabolites biosynthesis, transport and catabolism (89), transcription (78), post-translational modification, protein turnover, chaperones (71), carbohydrate transport and metabolism (61), and signal transduction mechanisms (59).
KEGG pathways were applied to analyze these DEGs. KEGG provided the integration of pathways, such as metabolic pathways, plant hormone signal transduction and plant-pathogen interaction pathways. 25,840 unigenes were mapped to 89 KEGG pathways based on KO terms for assignments (Fig. 8, Table S3). Using a probability of FDR ≤0.05, the top 20 items for KEGG pathways were presented in Figure 8. The most distributed pathways were phenylpropanoid biosynthesis (64), plant-pathogen interaction (27), glutathione metabolism (20).
Auxin-mediated genes in response to Cd stress
Through the analysis of DEGs, some auxin-mediated genes in responsive to cadmium stress were discovered (Table.3). Based on the transcriptome information, 141 significant DEGs were enriched in “plant hormone signal transduction” (Ko14488), most of which were participated in JA, auxin, GA, BR, and cytokinin signaling pathways (Table S4). Compared to CK, 14 related GO terms were identified, and two pathways were significantly enriched (p < 0.05) to further explore plant hormone biosynthesis, metabolism and signal transduction in Cd-treated groups (Table S4), including “plant hormone signal transduction” (GO: 0009734), and “zeatin biosynthesis” (GO: 0009851). For the GO term “auxin-activated signaling pathway”, we analyzed the expression of 28 auxin metabolism- and signaling-related genes and found that most of the genes were up-regulated in Cd (Table S5). In “response to brassinosteroid” genes encoding gibberellin-regulated protein 1-like, AP2/ERF and B3 domain-containing transcription factor RAV1 and steroid 5-alpha-reductase DET2 (LOC101248656, LOC101259430, and LOC101255051) were significantly down-regulated in Cd-treated groups, while another gene encoding gibberellin- regulated protein 11-like (LOC101260098) were up-regulated in Cd-treated groups (Table S5). Together, these results suggest that auxin metabolism and signal transduction, as well as BR affect the elongation of root and stem in Cd-treated groups.
Expression profile of DEGs encoding enzymes involved in lignin biosynthesis
Genes involved in the lignin biosynthetic pathway have been extensively studied in model plants and some fruit species. In the present study, seventeen immensely differentially expressed genes associated with the lignin biosynthetic pathway had been identified. Compared to control, 14 DEGs were up-regulated, three DEGs were down-regulated in Cd-treated tomato, which were respectively LOC101262425, LOC101262063 and LOC101265652 (Talbe. 4). 2 DEGs (LOC101265690 and LOC101265977) encoded caffeoyl-CoA o-methyltransferase (EC:2.1.1.104). 2 DEGs (LOC101249612 and LOC101246312) encoded peroxidase (EC:1.11.1.7). 5 DEGs (LOC101253619, LOC101262977, LOC101263266, LOC101262425 and LOC101249792) encoded laccase (EC:1.10.3.2). 3 DEGs (LOC101261765, LOC101247849 and LOC101262063) encoded 5-O-(4-coumaroyl)-D-quinate 3'-monooxygenase (EC:1.14.14.96). CTOMT1 encoded caffeic acid 3-O-methyltransferase (EC:2.1.1.68). Furthermore, there were also two genes (4CL and CCR1) that were well known in the lignin synthesis pathway, but their expressions were not significant. The results suggested that cadmium ion treatment activated lignin biosynthesis process.