3.1. TaEXPA9-A/B/D expression under abiotic stress and hormone treatments
TaEXPA9-A/B/D expression levels showed varying degrees of downregulation after PEG treatment, and these differences were significant compared with 0 h (P < 0.05). TaEXPA9-D expression decreased to 8.7% after 3 h of treatment. After 12 h of treatment, the TaEXPA9-A/B/D expression levels increased compared with 6 h, of which the TaEXPA9-A/B expression levels were significantly increased. After 24 h of treatment, TaEXPA9-A/B/D genes exhibited almost no expression (Figures 1A, D, and G). Under 4°C low temperature treatment, TaEXPA9-A gene expression gradually increased in the first 12 h, and the differences between the various time points were significant. TaEXPA9-A expression was downregulated at 24 h, though the difference compared with 0 h was not significant. TaEXPA9-B/D expression levels were upregulated at various time points after treatment compared with 0 h, and these differences were significant. TaEXPA9-A/B expression levels peaked at 12 h of treatment. Compared with 0 h, the expression levels of TaEXPA9-A and TaEXPA9-B were 163.6% and 136.7%, respectively (Figures 1A, D, and G).
Under MeJA treatment, the TaEXPA9-B/D expression levels gradually decreased as treatment duration increased, and the differences between the various treatment time points were significant (P < 0.05). The TaEXPA9-A gene was hardly expressed following MeJA treatment. Under the IAA, SA, and ABA treatments, TaEXPA9-A/B/D expression was almost completely inhibited (Figures 1B–C, E–F, H–I). In summary, TaEXPA9-A/B/D can be upregulated in response to 4°C low temperature stress.
3.2. TaEXPA9-A/B/D expression in different tissues and different growth stages
Quantitative RT-PCR was used to measure the expression of TaEXPA9-A/B/D genes in different tissues at different developmental stages. The expression levels of these three genes were the highest in the roots. The expression levels of TaEXPA9-B/D were the lowest in the stems, and the expression level of TaEXPA9-A was the lowest in the fruits. Hence, it can be deduced that these three genes may participate in stress responses in plants through similar signaling mechanisms (Figure 2).
3.3. Cloning and sequence analysis of TaEXPA9-A/B/D genes
In this study, the coding sequences (CDSs) of TaEXPA9-A/B/D in D2 seedlings at the two-leaf stage were cloned. The PCR amplification results showed that the length of the target genes was around 800 bp (Figure S1-A), which was consistent with the number of bases in the gene sequence. The CDS lengths of TaEXPA9-A and TaEXPA9-B/D were 807 and 801 bp, respectively. The similarity of the nucleotide sequences of these three genes was 97.23% (Figure S1-B-(a)). TaEXPA9-A/B/D contains three exons and two introns (Supply Figure S3-A). TaEXPA9-A encodes for 267 amino acids, and TaEXPA9-B/D encodes for 265 amino acids. The amino acid sequence similarity between the three genes was 96.14% (Figure S1-B-(b)). From this, it was deduced that the three sequences obtained by cloning were homologous genes. Protein structure prediction based on SMART showed that the TaEXPA9-A/B/D proteins contained DPDD_1 and Pollen_allerg_1 domains (Figure S1-B-(c/d)) and could induce cell wall loosening[9].
Table S1 shows the basic information of the TaEXPA9-A/B/D proteins. Their molecular weights ranged from 28.5–28.9 kDa and their isoelectric points (pIs) ranged from 9.40–9.61. These proteins are basic. The global average hydrophobicity (GRAVY) of these proteins was negative, showing that they are hydrophilic proteins. The instability indices (IIs) of these proteins were greater than 40, showing that these proteins are highly stable. The aliphatic indices (AIs) reflects the stability of the protein[42]. The aliphatic indices (AIs) of these proteins were greater than 60, showing that these proteins can adapt to the environment. In TaEXPA9-A/B/D proteins, alanine accounted for most of the amino acids (11–13%), while lysine contributed the lowest proportion (0.7–1.1%) (Table S2). The signal peptides in TaEXPA9-A and TaEXPA9-B/D are in positions 29–30 and 27–28, respectively (Figure S2-A). The three proteins are hydrophilic secretory proteins with transmembrane structures (Figures S2-B, C) and are located outside the cell membrane (Figure S2-D).The constructed three-dimensional structures of these proteins clearly show the catalytic domain (green region), cellulose binding domain (yellow and reg regions), and the signal peptide (dark blue region). There are slight differences in the signal peptide domain (1-1/2) between the TaEXPA9-A/D and TaEXPA9-B proteins and in the catalytic domain (2-1/2/3). These regions are labeled with a red rectangle in the figure (Figure S3-B). Comparison of EXPA9 proteins from eight plant species, which includes TaEXPA9-B/D proteins, and an EXPLA9 protein from one plant species (Figure S3-C) showed that all EXPA9s contain a common conserved sequence. This sequence contains three disulfide bonds formed by eight cysteine residues, a C-terminal cellulose binding domain consisting of six tryptophan residues, and one HFD (His-Phe-Asp) sequence in the center. MEGA7 was used to construct a phylogenetic tree. The results showed that TaEXPA9-A/B/D had the closest phylogenetic relationship with HvEXPA9 (Figure S3-D).
3.4. Subcellular localization of TaEXPA9-A/B/D
The CELLO v5.2 online prediction software predicted that the TaEXPA9-A/B/D proteins were located extracellularly (Figure S2-D). In order to validate the prediction results, we constructed 35S::eGFP, 35S::TaEXPA9-A-eGFP, 35S::TaEXPA9-B-eGFP, and 35S::TaEXPA9-D-eGFP vectors and transformed onion epidermal cells (Figure 3A). Fluorescence microscopy clearly showed that these three proteins were located on the cell wall (Figure 3B).
3.5. Phenotypic analysis of TaEXPA9-A/B/D-overexpressing A. thaliana plants
In order to validate the function of TaEXPA9-A/B/D, we separately constructed three 35S: TaEXPA9-A/B/D vectors, which were used for A. thaliana transformation (Figure 4A). Quantitative RT-PCR was used to measure TaEXPA9-A/B/D expression in A. thaliana (Figure 4B). Following that, overexpression vectors were transformed into A. thaliana. Resistance screening was used to identify overexpression plants, and the screening results are shown in Figure 4C. The WT and overexpression plants were cultured on vertical plates for 4 days before the hypocotyl length and number of lateral roots were measured. The mean hypocotyl length of the OE-A/D was 7.38 times that of the WT plants, while the mean hypocotyl length of OE-B was 6.75 times that of WT plants (Figures 4D and 5A). Microscopy was used to count the number of lateral roots at identical sites in the WT and overexpression A. thaliana roots. The results are shown in Figures 4E, 4F, and 5B. Compared with the WT plants, OE-A/B had 9–12 more lateral roots on average, and this difference was significant (P<0.05), and the OE-D had 6.3 more lateral roots on average, but this difference was not significant.
The number of root hairs and axial roots in A. thaliana was measured at 8 d. The results (Figure 4G and 5C) showed that the mean number of root hairs in the OE-A and OE-B/D were 1.49 and 1.69–1.75 times that of the WT, respectively, and these results were significant. The mean axial root length in the three overexpression A. thaliana strains was 0.30–0.48 cm longer than in the WT (Figures 4H and 5D), and these differences were significant. From this, it can be seen that the root system of the overexpression plants was more developed than the WT plants, and the overexpression plants possessed greater advantages than the WT plants under the same growth conditions. On 30 d of A. thaliana growth, the number of leaves, rosette diameter, and plant height of OE-A/B/D were significantly higher than in the WT (Figures 4I–K and 5E–G). The A. thaliana plants were grown until the fruiting stage. Figures 4L and 5H–I show that the number of pods and seed yield in the overexpression plants were significantly higher than in the WT plants. This confirmed that TaEXPA9-A/B/D promotes the growth of transgenic A. thaliana.
3.6. Phenotypic observation of TaEXPA9-A/B/D-overexpressing A. thaliana plants in response to cold stress
The D2 transcriptome analysis results showed that the expression of TaEXPA9-A/B/D was regulated by low temperature[37], and comparison and observation of the cold stress treatment on the overexpression plants and WT plants intuitively demonstrated the functions of these genes. Compared with the before treatment (Figure 6A), cold stress for 15 min caused leaf softening, wilting, and color deepening in the four types of plants (Figure 6B). After culturing in the plant culture room for 7 d, the growth of the plant leaves gradually recovered (Figure 6C). Phenotypic observation showed that the degree of recovery of OE-A/B/D were greater than the WT plants, but some plants died due to lower resistance. Recording and analysis of plant survival rates indicated that the survival rates of the OE-A/B/D were more than 70% (Table 3), being higher than in the WT. The difference in survival rate between groups was significantly associated with overexpression of TaEXPA9(x2 = 8.71, P = 0.033). Therefore, TaEXPA9-A/B/D had positive promoting effects on cold hardiness in transgenic A. thaliana.
3.7. Measurement of physiological markers in TaEXPA9-A/B/D-overexpressing A. thaliana plants under low temperature stress
The activities of antioxidant enzymes, SOD, POD, and CAT in the overexpression plants and WT plants at normal growth conditions and under 4°C treatment were maintained at high levels, and MDA level was low. The SOD activity of the OE-A/B fluctuated with the duration of stress, while the variation trend in SOD activity in the OE-D was stable. Under low temperature treatment for 3 h, SOD activity was 469.0 U/g, 443.6 U/g, and 317.6 U/g, respectively. At 12 h, SOD activity was 316.6 U/g, 483.9 U/g, and 314.3 U/g, respectively (Figure 7A). At low temperature treatment for 6 h, POD activity in the WT plant was at its peak of 224.36 U/g, which was 74.2%, 88.2%, and 54.7% that of the OE-A/B/D, respectively. At 24 h, POD activity in OE-B/D reached their peaks, which was 486.3% and 440.5% of WT plants, respectively (Figure 7B). At 24 h, the CAT content in the three overexpression plants peaked, being 1.60, 1.34, and 1.31 times that of the WT plants, respectively (Figure 7C).
The MDA content in OE-A/B/D at 24 h of treatment was 17.85, 10.71, and 11.82 nmol/g, respectively (Figure 7D), which was significantly lower than in the WT (P < 0.05). Comparison of various time points under normal growth conditions and low temperature treatment showed that the soluble protein, soluble sugar, and proline contents in the overexpression plants were all higher than in the WT plants. After 12 h of low temperature treatment, the soluble protein content in the overexpression plants peaked. By contrast, the soluble protein content in the WT plants was at its trough of 0.95 mg/g. The soluble protein content in the three overexpression plants was 426.3%, 362.1%, and 381.1% that in the WT plants, respectively (Figure 7E). As treatment duration increased before 12 h, the soluble sugar content in the WT plants first increased before decreasing, while the soluble sugar content in the overexpression plants gradually increased. After 24 h of treatment, the soluble sugar content in the overexpression plants decreased, but was still 220.3%, 242.5%, and 192.2% that of the WT plants, respectively (Figure 7F). At 6h, the proline content was high in the WT and overexpression plants, which was 11.02 µg/g in WT plants. The proline content in the OE-A/B/D were 227.7%, 279.9%, and 324.8% that of the WT plants, respectively (Figure 7G).