Overexpression of AmCBF1 enhances drought and cold stress tolerance, and improves photosynthesis in transgenic cotton

Cotton transformation

The full-length cDNA of the AmCBF1 (GenBank accession no. EU840990.1) gene was amplified and integrated into the pBI121 vector between the Xba I and BamH I restriction sites. The recombinant plasmid, pBI121-AmCBF1, was introduced into the Agrobacterium tumefaciens strain LBA4404 (Gu et al., 2013) to transform upland cotton variety R15 using the previously described protocol (Wang et al., 2017). After germination, the regenerated plants with true leaves and roots were transferred to soil.

Molecular identification

Genomic DNA was extracted from young leaves of transgenic cotton using the DNA secure plant kit (Tiangen, Beijing, China). PCR analysis was performed to detect AmCBF1 in transgenic plants using the specific primer pair AmCBF1-F and AmCBF1-R (Table S1) and the thermal cycling conditions (PCR thermal cycler, Gene Company, Shanghai, China) of 94 °C for 5 min; 35 cycles of 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 60 s; and 72 °C for 10 min. PCR products were separated using 1% (w/v) agarose gel electrophoresis and visualized with an ultraviolet spectrometer. Semi-quantitative reverse transcription PCR (RT-PCR) and quantitative real time PCR (qRT-PCR) were carried out to analyze the relative expression of AmCBF1 in R15 and transgenic cotton. Total RNA was extracted using the Plant RNA Kit (Yuanpinghao, Beijing, China) and cDNA was synthesized using the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix Kit (TransGen Biotech, Beijing, China) according to the manufacturer’s instructions. The specific primer pair AmCBF1-F and AmCBF1-R was used for RT-PCR. The PCR amplification protocol was as follows: 94 °C for 5 min; 30 cycles of 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 60 s; and 72 °C for 10 min. For qRT-PCR, the primers qA-F and qA-R were used. The amplifications were carried out at 95 °C for 1 min; 40 cycles of 95 °C for 15 s, and 60 °C for 30 s. Gene expression data was evaluated using the 2^−ΔΔCt method (Livak & Schmittgen, 2001). Small subunit ribosomal RNA (SSU) (GenBank accession number U42827) was used as the internal reference gene. The primer details are listed in Table S1. The lines showing the presence of transgenes were selected and propagated. T₄ generation seeds were used for further study.

To uncover the insertion site, whole-genome resequencing was performed by the LiBaijia company (Beijing, China). The high-quality clean data were mapped to the reference genome of upland cotton (https://cottonfgd.org/) (Zhu et al., 2017) using BWA software based on the resequencing results. To further confirm the exact insertion site of AmCBF1 in upland cotton, primers were designed for the pBI121-AmCBF1 plasmid and the upland cotton genome at both sides of the insertion site according to the results of whole-genome resequencing. The primers are listed in Table S1.

Phenotype analysis of transgenic cotton

To measure plant root length, shoot height, and root and shoot dry mass, whole R15 and transgenic cotton plants grown in the greenhouse were sampled randomly at the two-leaf stage. Averages were calculated for at least six separate individuals per line.

For seed germination analysis, transgenic lines L28, L30, and upland cotton variety R15 seeds were placed on sterile filter paper in a petri dish supplemented with water or mannitol (300 mM). The control and drought stress groups were cultured at 25 °C for 7 days, and the low temperature treatment group was placed in a climate chamber at 15 °C for 7 days. The germination rate was scored, and the radicle length of the seedlings was measured after 7 days. Thirty seeds were selected per line each time. The experiment was repeated four times.

Physiological analysis under drought and cold stress

The T₄ generation seeds of the homozygous transgenic cotton (L28 and L30) and R15 lines were sowed in pots containing a 1:1 mixture of peat and vermiculite in the greenhouse with 14 h light at 28 °C/10 h dark at 20 °C. When the two cotyledons of the seedlings fully expanded, the plants were divided into three groups: typical watering group (control), drought treatment group (300 mM mannitol), and cold treatment group (4 °C). The control plants were irrigated with water two times per week. For drought treatment, the plants were irrigated twice weekly with 300 mM mannitol for 30 days. For the cold stress, the plants were treated at 4 °C for 48 h. Following treatment, the physiological traits of the seedlings were measured as previously described: relative water content (RWC) of the leaves (Yue et al., 2012), electrolyte leakage from membranes in leaves (Gu et al., 2013), chlorophyll content (Arnon, 1949), and soluble sugar (Lv et al., 2007). Each cotton line under stress treatment included at least three replicates (pots).

Measurements of lower epidermal cell area, stomatal density, and photosynthesis

We washed the fourth true leaf from the R15 and transgenic cotton lines (L28 and L30) grown in the greenhouse at the five-leaf stage with water and treated them with 2.5% glutaraldehyde for 8 h at 4 °C. The lower epidermal cell area and stomatal density of the samples were observed for each line using a scanning electron microscope. The protocol was used as previously described (Beaulieu et al., 2008). The epidermal cell area was measured using Image J software (National Institutes of Health, Bethesda, MD, USA). Transgenic cotton and R15 lines were planted and divided into three groups as previously described. Each cotton line under each stress treatment consisted of at least three replicates. The photosynthesis of the fourth true expanded leaf in the control, drought, and cold treatment groups (15 °C for 48 h) was assessed using a portable infrared gas analyzer according to the instructions (Li-6400, Li-Cor Inc., Lincoln, NE, USA). Net photosynthesis, stomatal conductance, intercellular CO ₂concentration, and transpiration rate were measured on a clear warm day. The saturating photosynthetic photon flux density (PPFD) was maintained at 1,500 µmol m⁻² s⁻¹ emitted from a red-blue (665 and 470 nm, respectively) LED light source attached onto the leaf measurement chamber. The CO₂ concentration flowing into the leaf chamber was controlled at 400 µmol mol⁻¹. The cover area of the leaves was set as six cm². Leaf temperature was set at 25 °C, and the humidity was set to 50–70% for all measurements (Tricker et al., 2005).

Statistical analysis

All the assays above were repeated at least three times with three biological replicates. The means of three independent experiments were analyzed for significant differences using Duncan’s multiple range tests and SPSS software (IBM SPSS statistics version 21, IBM, Armonk, NY, USA). Results were considered statistically significant when P < 0.05.

Generation and molecular identification of transgenic cotton overexpressing AmCBF1

The T-DNA structure of the pBI121-AmCBF1 and the location of related primers are shown in Fig. 1A. The representative images during the transformation process of cotton are shown in Fig. 1B. The PCR results indicated that 11 independent transgenic cotton lines were obtained after Agrobacterium-mediated genetic transformation (Fig. 1C). Two dominant homozygous lines, L28 and L30, were selected based on PCR results for further analysis. The RT-PCR results showed that the AmCBF1 gene was expressed in both cotton lines but not in R15. qRT-PCR showed that the expression level of AmCBF1 in L30 was 1.73-times higher than that in L28 (Fig. 1D). Comparing whole-genome resequencing data with the upland cotton reference genome indicated that there were 10 and 22 pairs of matched reads separately on chromosomes A05 and A12, respectively, in L28, and 24 and 11 pairs of matched reads separately on chromosomes A12 and A13, respectively, in L30. The whole-genome resequencing showed that the insertion sites of L28 and L30 were located at 8998563–8998590 on chromosome A05 and 4379624–4379637 in chromosome A13, respectively (Fig. S1). Three pairs of different primers were designed to verify the left and the right borders of the insertion sites on chromosome A05 and A13 separately. The PCR results further confirmed the insertion site that was determined using genome resequencing (Fig. 1E). Intriguingly, the resequencing data indicated that both L28 and L30 contained some matched reads on chromosome A12, and the insertion site was almost always located at 86125018–86125479. Three pairs of different primers were designed to evaluate the insertion site on chromosome A12. The results showed that the specific target fragment amplified in L28 or L30 could also be obtained in R15 (Fig. S2). Further sequencing analysis revealed an endogenous GhDREB (GenBank accession No. XM_041083773.1) sequence existed in this region. The homology between the mRNA sequence of this GhDREB with AmCBF1 was 57.37%, and the core sequence of this GhDREB with AmCBF1 was 77.08% (Fig. S3). The amino acid sequence homology between this GhDREB and AmCBF1 was 58.56% (Fig. S4). These results proved that there was no T-DNA insertion in chromosome A12 of L28 and L30, and they were both single-copy insertion.

AmCBF1 participation in plant growth regulation

Phenotypes were examined under normal growth conditions when the first two true leaves of cotton had fully expanded (Fig. 2A). The root length of L28 (about 12.63 cm) was significantly longer than that of L30 (10.58 cm). The difference in root length of L28 and L30 was not significant in comparison to R15 (11.92 cm). The average plant height of L28 (13.51 cm) and L30 (12.41 cm) was significantly shorter than that of R15 (17.50 cm) (Fig. 2B₁). The root to shoot ratio of the transgenic seedlings (more than 0.10) was significantly higher than that of R15 (0.08) (Fig. 2B₂).

Seed germination and root growth under drought and cold stress

Seed germination under different treatments was observed and recorded. In the control group, there were no differences in the germination rates among the R15 and two transgenic lines 7 d after sowing at 25 °C. Germination was delayed by 1 or 2 days across all seeds under drought or 15 °C conditions. The germination rate of the R15 seeds was 78.33% in drought and 2.50% in 15 °C treatments, which was significantly lower than that of L28 (97.50% in drought, 41.67% in cold) and L30 (100% in drought, 71.67% in cold) (Figs. 2C and 2D₁). There was no significant difference in the root length of R15 and transgenic cotton 7 d after sowing at 25 °C. However, after treatment with 300 mM mannitol, the root length of R15 was about 2.08 cm, compared with 2.60−2.68cm in the transgenic lines. Similarly, the root length of R15 was 0.20 cm compared with 0.42 cm of L28 and 0.49 cm of L30 following cold treatment. The root lengths of these two transgenic lines were always significantly longer than R15 under drought or cold stress (Figs. 2C and 2D₂). Although root growth is expected to be inhibited at low temperatures, it was less inhibited in the transgenic plants than in R15. These results suggested that overexpression of AmCBF1 gene in cotton can improve drought and cold tolerance during the seed germination stage.

Physiological analysis under drought and cold stress

To determine whether AmCBF1 increased the resistance of transgenic cotton plants to drought or cold stress, multiple physiological indicators were measured before and after drought and low temperature treatments. The RWC in transgenic and R15 plants was more than 80% under normal conditions, and always significantly higher in transgenic lines than in the wild types. Compared with the control group, the RWC of transgenic cotton and R15 decreased under drought or cold stress. The RWC in L28 (about 80.23%) and L30 (about 84.08%) were significantly higher than in R15 (72.55%) after drought treatment. Under cold treatment, the RWC in the two transgenic lines (over 70%) were significantly higher than in R15 (65.52%) (Fig. 3A). Electrolytic leakage in the leaves of transgenic cotton and R15 were around 25% under normal conditions, but were up to 39.48% in R15, 29.71% in L28, and 28.72% in L30 under drought stress. Under cold treatment, the electrolyte leakage reached 56.15% in R15 and less than 30% in the two transgenic lines. These two transgenic cotton lines showed notably less electrolyte leakage compared to R15 under drought or cold stress (Fig. 3B). AmCBF1 appeared to confer protection against membrane leakage in the transgenic lines. The chlorophyll content in transgenic cotton was always significantly higher than in R15. The average chlorophyll content decreased after 300 mM mannitol or low temperature treatment. It was reduced by 22.25% (R15), 21.86% (L28), and 18.70% (L30) after drought treatment, and 31.21% (R15), 27.28% (L28), and 20.18% (L30) after cold treatment. The decreasing trend of chlorophyll content was more moderate in transgenic cotton than in the control group (Fig. 3C). The content of soluble sugars in the two transgenic lines was significantly higher than in the wild type under normal conditions (1.94 times higher in L28 and 2.22 times higher in L30). L28 was 1.60 times higher and L30 was 2.37 times higher than R15 under drought stress. L28 was 2.10 times higher and L30 was 1.73 higher than R15 under low temperature stress. Drought and low temperature stress significantly increased the soluble sugar content in both R15 and the transgenic lines compared with the controls (Fig. 3D). These results further suggested that AmCBF1 overexpression conferred drought and cold tolerance to cotton.

Measuring low epidermal cell area, stomatal density, and photosynthesis in cotton seedlings

The epidermal cells were usually irregular or polygonal in R15, but they were mostly rectangular and arranged regularly and tightly in the transgenic lines (Fig. 4A). The transgenic lines also had a markedly smaller area of epidermal cells and higher stomatal density on leaves than R15 (Fig. 4B). Net photosynthesis, stomatal conductance, intercellular CO₂ concentration, and transpiration rate in the transgenic cotton lines were all significantly higher than R15 under normal conditions (Fig. 4C). When compared with those under normal conditions, net photosynthesis, stomatal conductance, and transpiration rate were dramatically inhibited by drought and cold treatments, but were all significantly higher in the transgenic lines than in R15. Intercellular CO₂ concentration was higher in transgenic cotton than in R15 before and after drought treatment, but there was no difference under cold stress (Fig. 4C). AmCBF1 overexpression improved leave photosynthesis of cotton seedlings, which may be due to AmCBF1 regulating epidermal cell differentiation and formation and improving photosynthesis-related parameters, such as stomatal density.