PLB morphological characterization

The plant materials from the four growth stages (described above) were observed and photographed under a stereo microscope (Nikon, SMZ745T, Japan) to study the morphological characters.

The samples from the four growth stages (described above) were fixed in formalin acetic acid-alcohol solution (FAA; 70% ethyl alcohol: glacial acetic acid: 37% formaldehyde; 18:1:1) for a week. Fixed samples were dehydrated in a series of alcohols (70, 85, 95, 100, and 100%; v/v) for 1 h. Then, the samples were immersed in an ethanol-xylene mixture (2:1, 1:1, and 1:2; v/v) for 1 h, and xylene solution for 2 h. Paraffin was added to the solution until saturated at 36 °C overnight. Next, the temperature was gradually raised to 58 °C, and xylene was replaced by paraffin. Finally, the materials were embedded by paraplast after soaking in pure paraffin for 3 h. The wax blocks were sliced into 8 μm segments with a microtome (KEDEE, China). The dye used was Ehrlich’s haematoxylin (Biosharp, China), 1% safranin O (Solarbio, China), and 0.5% fast green dyes (Solarbio, China). The sections were observed and photographed under a biological microscope (Nikon, E200, Japan).

The determination of plant hormones was performed by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Tissue culture materials from each of the four growth stages were taken as experimental materials. Plant samples (50 mg) were ground into a powder in liquid nitrogen and added to 10 times the volume of acetonitrile solution. The material was extracted overnight at − 4 °C. After centrifugation at 12,000×g for 5 min, the supernatant was collected. To extract the hormone, 5 times the volume of acetonitrile solution was added to the precipitate, and the supernatant was collected and merged with the previous supernatant. After adding 25 mg CNW C18 QuEChers packing, the mixture was vortexed vigorously for 30s and centrifuged at 10,000×g for 5 min. The supernatant was collected and dried with nitrogen stream. The residues were redissolved in 200 μL methyl alcohol and filtered through 0.22 μm organic phase membrane. The mass spectrometer used is the Qtrap6500 (Agilent, America).

The results of the samples in each growth stage were represented by the average value of three biological replicates. The treatment data were subjected to analysis of variance (ANOVA) using IBM SPSS Statistics version 20. Duncan’s multiple range test at P value < 0.05 was performed with stage A as the control.

Column Plant RNAout2.0 (Tiandz Inc., Beijing, China) was used for total RNA extraction performed according to the manufacturer’s instructions. Approximately 100 mg of each sample was used to extract RNA. The RNA was assessed using agarose gel electrophoresis, Nanodrop One (Nanodrop Technologies Inc., DE, USA), and Agilent 2100 (Agilent Technologies Inc., CA, USA) to confirm the purity, concentration, and integrity, respectively. The 260/280 nm ratios and 260/230 nm ratios of 1.8–2.2 and 1.6–2.2, respectively, from the Nanodrop were regarded as pure. Next, the RNA library was constructed, and sequencing was performed by Genepioneer Technologies Corporation (Nanjing, China). HiSeq4000 platform (Illumina Inc.) was used for high-throughput sequencing with a read length of PE150.

Transcriptome sequencing was performed on the samples from the four growth stages (described above), with three biological replicates for each stage. Raw data produced by sequencing were processed by removing adapters as well as filtering low quality reads with over 10% high unknown base (N) reads to obtain high quality clean data. Phred quality score Q20 and Q30 and GC-content of clean reads were calculated. Clean reads were assembled to finally obtain the unigene library of P. SCBG Huihuang90. The quality of transcriptome sequencing libraries was evaluated from three different perspectives: (1) examining the distribution of inserted fragments on unigene to evaluate the randomness of mRNA fragmentation and the degradation of mRNA; (2) drawing the length distribution map of the inserted fragment to evaluate the dispersion degree of the length of the inserted fragment; (3) evaluating whether the library capacity and the mapped reads compared to the unigene library were sufficient by drawing saturation map. All sequence data were uploaded into the BioProject database hosted by the National Center for Biotechnology Information (NCBI) under the BioProject PRJNA684752. Software BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi) was used for functional annotations by comparing the information of sequence or amino acid sequence of unigenes to 7 databases: NR, Swiss-Prot, KEGG, COG, KOG, GO, and Pfam.

Bowtie (http://bowtie-bio.sourceforge.net/index.shtml) was used to compare the sequenced reads with unigene library. Fragments per kilobase per million mapped reads (FPKM) value was used to estimate the expression abundance of unigene. DEGs between libraries were identified by DESeq2 (http://www.bioconductor.org/packages/release/bioc/html/DESeq.html). Fold change represents the ratio of expression quantity between two samples, and the Benjamini-Hochberg approach was used to adjust the P values for controlling the FDR. Unigenes with FDR < 0.05 and an absolute value log2 (Fold change) ≥ 1 were considered differentially expressed. KEGG enrichment of DEGs was measured by enrichment factor, q value, and the number of genes enriched in the corresponding pathway.

To validate the results of RNA-seq data, quantitative real-time PCR analysis was used to detect the expression levels of the 12 candidate genes. Primers were designed based on the Primer-BLAST [71] and listed in Supplemental Table S6. ACT2 (TRINITY_DN57670_c1_g1) was selected as the reference gene for the normalization of the data. RNA of each growth stage was reverse transcribed using the One-Step gDNA Removal and cDNA Synthesis SuperMix kit (Trans, Beijing, China) according to the instruction manual, and cDNA of approximately 600 ng/μL was obtained. cDNA was diluted three times for subsequent experiments. 1 μL cDNA mixed with 0.8 μL primer pair (10 μM), 10 μL 2× Green qPCR SuperMix (Trans, Beijing, China), and 8.2 μL ddH2O. The mixture was used to carry out qRT-PCR detection on LightCycler 480 System (Roche Diagnostics, Germany). The amplification program was performed as follows: 30 s at 94 °C for initial denaturation, 40 cycles of 5 s at 94 °C for denaturation, 15 s at 57 °C for annealing, 10 s at 72 °C for elongation, followed by melting curve program for melting curve analysis. Each sample was processed as three biological replicates and three technical replicates. The relative expression was calculated by the 2^-ΔΔCt method [72].