Plant growth conditions and drought treatment
For this study, we selected the Lanzhou lily and the Tresor lily for our comparative experiment. Lanzhou lily bulbs were purchased in the West Orchard of Lanzhou City, Gansu Province, and Tresor were purchased in Xiwang Flower Seed Industry Co., Ltd. A voucher specimen of Lanzhou lily was deposited at the Northwest Institute of Eco-environment and Resource, Chinese Academy of Sciences (number LZ 00321). The experiment was carried out in Gaolan Research Station under the auspice of the Chinese Academy of Sciences (CAS) (Gaolan County, Lanzhou City, Gansu Province, China; 36°13″N 103°47″E). This area is a typical Semi-arid Loess Hilly and gully area with dry and rainless climate. It is a warm temperate semi-arid continental monsoon climate with an altitude of about 2100 m, an annual average temperature of 6.8 ℃.
Lanzhou lily and Tresor were grown in greenhouses, with ventilation on both sides of the shed, which only prevented rain. The experiment was carried out in the way of barrel planting. 7 kg of substrate (soil: peat soil = 3:1) was packed in the same plastic barrel (specification 24 cm × 28 cm). Lily bulbs with the same size were selected and transferred into the barrel, and the bulbs were buried about 2-3 cm under the substrate.
Drought stress treatment: the experiment adopted completely random design, determined the soil water content by drying method, and determined the intensity of drought stress according to the soil water content and stress days. There were three treatments, namely control group (c), moderate stress (MS) and severe stress (s). The actual soil moisture content was (12±1.23)%, (8±1.3)% and (6±0.7)%, respectively. The experiment started in March 2019, and began to harvest at the end of the whole growth cycle. Five plants were planted per barrel, 90 barrels for each variety, and 15 repetitions for each treatment. Before the drought treatment, all the treatments should be watered with enough water to ensure the emergence of seed balls. After the emergence rate reaches 99%, the water control treatment should be carried out after one-time quantitative irrigation. At 50 days of drought stress, the middle healthy leaves were taken to measure the indexes.
Measurement of physiological indexes
The leaf relative water content was determined according to the method established by Barrs and Weatherley (1962)[36]. Chlorophyll concentrations (determined using the soil plant analysis development (SPAD) value) were measured using a SPAD 502 Plus Chlorophyll Meter (Konica Minolta, Inc., Sakai City, Osaka, Japan). Stomatal apertures were observed using light microscope (Leica DM750) following 50 days of the drought stress treatments. Stomata were counted in 5 randomly selected microscope fields, and each sample was repeated three times.
Scanning electron microscopy (SEM)
Lanzhou lily and Tresor of leaf were cut along the leaf vein from the middle, divided into positive and negative sides. The samples were cut at a distance of 0.5 cm from the middle leaf vein, fixed in a test tube with glutaraldehyde and then dried in an oven at 40 ℃, and cut into 3mm x 3mm placed in the coater to spray gold coating. The morphology was observed under the scanning electron microscope (S3000-N, Hitachi, Tokyo, Japan).
Wax extraction
We extracted the wax according to the previous method[37]. Five lily plants with the same growth condition were selected from the control group and the drought stress group. The leaves in the middle of two kinds of lilies were separated with scissors, and the leaf area was determined rapidly. The leaves were immersed in 30 ml chloroform at room temperature immediately, shaken gently for 1 min, and then the leaves were taken out. Then 20 μL C24 alkanes were added into the extraction solution as internal standard. After the chloroform in the sample bottle was dried with nitrogen, 40 μL pyridine was added to dissolve the wax, and then the derivative BSTFA of the same volume was added. The derivative was performed in a metal bath at 70 ℃ for 30 s. Dry the derivative reagent with nitrogen, dissolve the product in 1 ml of chloroform, and transfer it into a sample bottle for GC-MS analysis.
Wax components analysis
The wax content was determined by gas chromatograph (AOC-20i, GC-2010, E). Using GC-MS to detect the waxy components in the control group to get the ion peaks of various waxy compounds, and then through the retrieval of the mass spectrometry database to identify various compounds to get the standard sample map. After GC-FID was used to detect each sample, the ion peak of wax component was deduced according to the reference diagram, and then the ion peak area was obtained by integrating the ion peak with laboratory software, then the various compounds were quantitatively analyzed according to the internal standard peak area and the content of C24 added. GC-MS and GC-FID: injection volume 2 µL, injection temperature 280 ℃, split ratio 5:1, detector temperature 320 ℃.
Phylogenetic tree analysis of KCS gene
We used NCBI (http://www.ncbi.nlm.nih.gov/blast/) biological software for amino acid multi sequence alignment analysis to find homologous genes in Lanzhou lily and other species, and used Mega 5.0 to construct phylogenetic tree of amino acid sequences by neighbor joining method.
RNA extraction and construction of cDNA library
Total RNA was extracted from frozen leaves of 9 plants of three independent biological replicates with rnaprep plant Kit (tlangen biotechnology Beijing) according to the instructions of the kit. The purity, concentration and integrity of RNA samples were detected by nanodrop and Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). Adaptor sequences and low quality reads were removed, mRNA was enriched by oligo (dT) beads and then cleaved into short fragments using fragmentation buffer. Using mRNA as template, the first cDNA strand was synthesized by random hexamers, the second cDNA was synthesized by buffer, dNTPs, RNase H and DNA polymerase I, the cDNA was purified by AMPureXPbeads, the purified double cDNA was repaired at the end, followed by the addition of poly-(A) tails and ligation to Illumina sequencing adapters, and then the fragment size was selected by AMPure XPbeads. Finally, the cDNA library was obtained by PCR enrichment. After the sample was tested to be qualified, the library was constructed. High quality clean data of 211.64 GB was obtained, and the percentage of base Q30 reached 90.53% (Table 3). The clean data were combined in order to obtain the single gene library of the species. Through the assembly of Trinity software, 324081 transcripts and 142191 single genes were obtained (Table 4).
Differentially expressed genes (DEGs) analysis
BowTie software was used to compare sequencing reads and UniGene library sequences, while RSEM was used to estimate their combined expression levels[38]. Fragments Per Kilobase of transcript per Million mapped reads (FPKM) were used to represent the expression abundance of the corresponding UniGene sequences. The log2 (fold change) ≥ 1 and the false discovery rate (FDR) < 0.05 thresholds were used to obtain statistically significant gene expression differences.
Statistical analyses
The data were analyzed by Excel and spass16.0 software with mean ± standard error. One way ANOVA was used to compare the differences. Duncan's method was used to test the difference of wax content between varieties and treatments. We defined significance at P < 0.05.