Identification and expression analysis of the DREB transcription factor family in pineapple (Ananas comosus (L.) Merr.)

Identification of DREB family members in pineapple

Dehydration responsive element-binding amino acid sequences from Oryza sativa and Arabidopsis thaliana were obtained from the Rice Genome Annotation Project (RGAP, http://rice.plantbiology.msu.edu/index.shtml) (Kawahara et al., 2013) and The Arabidopsis Information Resource (TAIR, http://www.arabidopsis.org) (Berardini et al., 2015), respectively. The DREB sequences from Arabidopsis were used as search queries in BLAST-P against the pineapple genome. The AP2 (PF00847) domain was downloaded and used as a query to perform a HMMER search with default parameters (https://www.ebi.ac.uk/Tools/hmmer/search/phmmer). HMMER is a software package that uses profile hidden Markov Models to identify conserved domains (Ming et al., 2015). Redundant sequences were eliminated and the Simple Modular Architecture Research Tool (SMART, http://smart.embl-heidelberg.de/) (Letunic & Bork, 2018) was used to verify the existence and completeness of the core domain within the identified sequences. The sequences that met these criteria were used for phylogenetic analysis.

Protein characteristics and chromosomal localization

For each of the putative AcoDREB genes, the gene length, amino acid number, coding sequence (CDS) length, and chromosome position were collected from the Pineapple Genomics Database (PGD, http://pineapple.angiosperms.org/pineapple/html/index.html) (Xu et al., 2018). The molecular weights and isoelectric points of the putative proteins were predicted using the ExPASy proteomics server (http://expasy.org/) (Gasteiger et al., 2003). Based on the start positions of the genes and the lengths of the corresponding chromosomes, MapChart (Voorrips, 2002) was used to visualize the 20 AcoDREB genes that were mapped onto the 25 pineapple chromosomes and scaffold sequences.

Cis-element analysis of AcoDREB gene promoters

The 2 kb upstream sequences of the AcoDREB genes were retrieved from the Pineapple Genomics Database and submitted to Plant Cis-Acting Regulatory Element (PlantCARE, http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al., 2002) to detect the presence of the following six regulatory elements (Sazegari, Niazi & Ahmadi, 2015): abscisic acid (ABA)-responsive elements (ABREs; ACGTG/TC), which are involved in ABA responsiveness (Yamaguchi-Shinozaki & Shinozaki, 1993); dehydration-responsive elements (DREs; A/GCCGAC), which are involved in plant responses to dehydration, low temperature, and salt stress (Narusaka et al., 2003); low temperature-responsive elements (LTREs; CCGAA), which are involved in low-temperature responses (Roy Choudhury et al., 2008); TC-rich repeats (G/ATTCTCT), which are involved in defense and stress responses (Diaz-De-Leon, Klotz & Lagrimini, 1993); W-boxes (TGACC/T), which are the binding site of WRKY TFs in defense responses (Jiang et al., 2017); and MBS (TAACTG), or MYB binding sites, which are involved in drought-inducibility (Urao et al., 1993).

Sequence alignment and phylogenetic analysis

The CDS of the AcoDREB genes were obtained from the Pineapple Genomics Database and imported into DNAMAN Version 9 for sequence alignment (Wang, 2016). The phylogenetic tree was constructed in IQ tree using the maximum likelihood method (Chernomor, Von Haeseler & Minh, 2016; Nguyen et al., 2015). For this analysis, the parameters were set to default, except for the ultrafast bootstrap option, which was set to n = 1,000 (Hoang et al., 2018), after performing multiple sequence alignments in MUSCLE 3.7 (Edgar, 2004) using default parameters. To validate the maximum likelihood results, the neighbor-joining method was used to construct a tree using MEGA7 (Kumar, Stecher & Tamura, 2016).

Gene structure analysis and conserved motif identification

The DREB gene structures, including the numbers and positions of exons and introns, were determined using the Gene Structure Display Server (GSDS, http://gsds.cbi.pku.edu.cn/) (Guo et al., 2007). Multiple EM for Motif Elicitation (MEME, http://meme-suite.org/tools/meme) was used to analyze the amino acid sequences of the 20 AcoDREBs; the maximum number of motifs was set to 10, and default parameters were used (Bailey et al., 2009).

Plant material and growth conditions

The pineapple (Ananas comosus) variety MD2 was provided by the Qin Lab (Haixia Institute of Science and Technology, Fujian Agriculture and Forestry University, Fujian, China) (Priyadarshani et al., 2018). Plants were grown on a soil mixture containing 2:1 (v/v) peat moss:perlite, in plastic pots in a greenhouse under the following conditions: 30 °C, 60–70 μmol photons m⁻¹ s⁻¹ light intensity, 70% humidity, and a 16-h light/8-h dark photoperiod.

RNA-Seq of different pineapple tissues

We used an RNA extraction kit (Omega Bio-Tek, Shanghai, China) to extract total RNA from the following tissues: calyx, gynoecium, ovule, petal and stamen. The tissues were collected according to previously described methods (Chen et al., 2017). The NEBNext Ultra RNA Library Prep Kit for Illumina was used to prepare libraries prior to sequencing. RNA-Seq data for root, leaf, leaf base, leaf tip, flower and fruit at different development stages were collected from the Pineapple Genomics Database (Ming et al., 2015). Using TopHat v2.1.1 (Trapnell et al., 2012) with default parameters, the trimmed paired-end reads of all samples were aligned to the pineapple genome. Cufflinks v2.2.1 and Cuffdiff v2.2.1 were used to estimate the Fragments Per Kilobase of exon model per Million mapped values. The heatmap showing the AcoDREB gene expression profile was generated using the pheatmap package in R (Galili et al., 2018).

Stress treatments

One-month-old plants in rooting medium were used as the planting material for the stress treatment analyses. Uniform tissue-cultured seedlings were obtained from the Qin Lab (Priyadarshani et al., 2018). Seedlings were subjected to the following stress treatments: low temperature (4 °C), high temperature (45 °C), drought (350 mM mannitol), and high salt (150 mM NaCl). Root and leaf tissues were collected at 6, 12, 24 and 48 h after treatment. Seedlings that were not subjected to any of the stress treatments were used as controls. The collected samples were immediately stored in liquid nitrogen prior to total RNA extraction (Rahman et al., 2017).

Quantitative real-time PCR and data analysis

Total RNA was extracted using the Plant RNA Kit (Omega Bio-Tek, Shanghai, China) according to the manufacturer’s instructions. The RNA concentrations ranged from 100 to 500 ng/μl, and the OD₂₆₀/OD₂₈₀ ratios ranged from 1.8 to 2.0. According to the supplier’s instructions for AMV reverse transcriptase (Takara Bio, Beijing, China), 1 μg of purified total RNA was reverse transcribed into cDNA in a total reaction volume of 20 μl (Cai et al., 2019). To quantify the relative transcript levels of selected DREB genes, real-time PCR was performed with gene-specific primers on the Bio-Rad Real-time PCR system (Foster City, CA, USA) according to the manufacturer’s instructions. The gene-specific primers used for this analysis are listed in Table S1. The PCR program used the following conditions: 95 °C for 30 s, 40 cycles of 95 °C for 5 s and 60 °C for 34 s and 95 °C for 15 s. For all tested genes, three technical replicates and at least three independent biological replicates were used (Cai et al., 2017; Zhang et al., 2018). Relative expression was calculated using the 2^−ΔΔCt method (Century, Reuber & Ratcliffe, 2008). Data were analyzed using one-way analysis of variance (ANOVA). Significant differences between treatments and controls are indicated by asterisks (* indicates a p-value < 0.05 and ** indicates a p-value < 0.01) (Table S2).

Genome-wide identification and chromosomal locations of pineapple DREB genes

Using Arabidopsis DREB amino acid sequences as search queries in BLAST, 20 DREB amino acid sequences were obtained from the pineapple proteome. The corresponding genes were named AcoDREB1 to AcoDREB20 (Table S3), and the amino acid sequences are listed in Table S4. Table 1 lists the following information for the 20 genes: gene name, gene ID, nucleotide and amino acid lengths, and the predicted isoelectric point (pI) and molecular weight (Mw) of the encoded protein. The protein lengths ranged from 149 (AcoDREB13) to 463 (AcoDREB20) amino acids, and the CDS lengths ranged from 450 (AcoDREB13) to 1392 (AcoDREB20) bp. The predicted protein molecular weights ranged from 16316.44 (AcoDREB13) to 49311.65 (AcoDREB20) Da, and the predicted isoelectric points ranged from 4.71 (AcoDREB10) to 9.68 (AcoDREB07) (Table S5). The 20 AcoDREB genes mapped to 14 pineapple chromosomes (Fig. 1), with three genes on Chr2 and two genes each on Chr3, Chr5, Chr6 and Chr17. Nine other chromosomes each contained one AcoDREB gene.

Multiple sequence alignment and phylogenetic analysis of the DREB family

Multiple sequence alignment of the AcoDREB AP2 domains indicated that the domain was highly conserved among the 20 AcoDREBs, and that it displayed characteristics typical of other DREB proteins (Fig. 2). Beyond the conserved YRG and RAYD motifs, all 20 AP2 domain sequences contained a Val residue at position 14 (Val14), and 11 of them had a Glu residue at position 19 (Glu19). Val14 is more important than Glu19 for the binding of DREB to the DRE cis-acting elements (Sakuma et al., 2002).

To determine the phylogenetic relationships between the DREB family members, we constructed a multi-species phylogenetic tree using the full-length amino acid sequences of DREBs from pineapple, Arabidopsis (Table S6) and rice (Table S7). In Fig. 3, AT3G57600 and AT2G40220 (red frame) belong to the Arabidopsis subgroups A-2 and A-3, respectively. Because none of the pineapple DREB genes were homologous to the A-3 subgroup, we divided the AcoDREBs into five subgroups, I to V (Fig. 3). Group I included AcoDREB01, 02 and 03, group II included AcoDREB04, 05, 06 and 19, group III included AcoDREB07, 08, 09 and 10, group IV included AcoDREB11, 12, 13, 14 and 15, and group V included AcoDREB16, 17, 18 and 20.

Stress-related cis-elements in AcoDREB promoters

Because of the potential involvement of AcoDREB genes in stress responses, we investigated the distribution of stress-related conserved cis-elements in their promoter regions (2 kb region upstream of the transcription start site) using PlantCARE (Table S8). The data for six abiotic stress response elements, ABRE, DRE, LTRE, TC-rich repeat, MBS and W-box, are shown in Fig. 4. All of the AcoDREB genes possessed at least one kind of cis-acting regulatory element, indicating that AcoDREB expression is associated with abiotic stress. Nine AcoDREBs had one or more LTREs, which are associated with the response to low-temperature conditions. Sixteen AcoDREBs contained between one and eight ABA-responsive elements, and only AcoDREB09, 12 and 17 had the TC-rich repeat element. Seven AcoDREBs had the MBS element, while W-boxes and DREs both occurred in ten AcoDREBs. Overall, the results of the cis-element analysis indicate that AcoDREB genes can respond to different kinds of abiotic stresses.

AcoDREB gene structure and conserved motifs in the encoded proteins

Structural diversity is very common among duplicated genes, and may result in the evolution of functionally distinct paralogs. To analyze the AcoDREB gene structures, exon and intron numbers and positions were determined by comparing the full-length cDNA sequences to the corresponding genomic DNA sequences (Fig. 5). Seventy five percent of the AcoDREB genes (15/20) lacked introns. Four genes (AcoDREB18, 04, 19 and 13) had one intron each, and AcoDREB05 had three introns. Interestingly, the members of group II differed in terms of exon and intron number as well as UTR length, which suggests that these four paralogs may have different roles in pineapple growth and development.

As shown in Fig. 6, the distribution of the motifs among AcoDREB proteins was relatively conserved. Motifs 1, 2 and 3 were present in all genes, but the motifs in different subgroups indicated some degree of divergence among them. For example, the three members in subgroup I contained motifs 4, 5 and 9 in addition to motifs 1, 2 and 3. Motif 7 was only present in two of the subgroup III proteins (AcoDREB07 and AcoDREB08), and motif 4 was only present in AcoDREB05 of subgroup II. Generally, members within the same subgroup had similar motif compositions, indicating that they may perform similar functions (Fig. S1).

AcoDREB gene expression profiles in different tissues at different developmental stages

The different stages of the reproductive organs were defined according to previous studies (Azam et al., 2018; Su et al., 2017). We used transcriptome sequencing data to analyze the expression patterns of the 20 AcoDREB genes in nine different tissues: root, leaf, flower, fruit, gynoecium, stamen, petal, calyx and ovule (Fig. 7; Table S9). We also used quantitative real-time PCR (qRT-PCR) to verify the results of the RNA-seq. All AcoDREB genes, except four that had low levels of expression (AcoDREB04, 07, 08 and 13), were selected for qRT-PCR analysis in seven tissues. The results obtained were consistent with the RNA-Seq expression data of these genes (Fig. 8; Table S10).

Clustering analysis of the expression patterns of the 20 genes divided them into four clusters (Fig. 7). Of the six genes in cluster I, four (AcoDREB05, 16, 17 and 20) were highly expressed in all tissues, indicating that they may have important roles throughout plant growth. The expression level of AcoDREB09 was lower in stamens than in other tissues, and AcoDREB19 had the lowest expression in roots, suggesting that these particular cluster I genes may not be critical for the development of these respective tissues. The six genes in cluster III (AcoDREB02, 04, 07, 08, 10 and 13) had very low expression levels in all tissues, suggesting that these genes might only be expressed under specific conditions. Most of the genes in clusters II and IV had tissue- or stage-specific expression patterns. For example, AcoDREB01 and AcoDREB15 had higher expression in calyxes, suggesting that they may have a positive role in floral organ development. The higher expression of AcoDREB06 in stage 6 stamens suggests a potential link to stamen maturity. AcoDREB18 was highly expressed during stamen development. AcoDREB11 was expressed in the ovule, stamen and gynoecium tissues, suggesting this gene may function widely during gametophyte development. AcoDREB03 was highly expressed in the root, calyx, petal, and gynoecium.

AcoDREB gene expression under abiotic stress

We analyzed AcoDREB gene expression under various abiotic stress conditions, including salt, drought, cold, and heat. Specifically, we examined the expression patterns of eight AcoDREB genes (AcoDREB01, 03, 06, 09, 11, 14, 18 and 19) in the MD2 variety of pineapple using qRT-PCR with three biological and three technical replicates (Fig. 9; Table S2). Under all stress conditions, the relative transcript levels of the AcoDREB genes fluctuated during the 48-h analysis period.

We subjected pineapple plants to salt stress by treating them with 150 mM NaCl. The expression of all eight genes increased rapidly in the roots and reached a maximum level 12 h after the start of treatment. In shoots, five of the genes had highest expression levels at 12 h, and two genes had highest expression levels at 6 h. AcoDREB06 expression in shoots decreased after salt treatment. The differential responses of the AcoDREB genes after NaCl treatment suggest that they have distinct roles in salt stress response (Figs. 9A–9H).

To analyze the response to drought stress, we treated plants with 350 mM mannitol. In the shoots, six genes (AcoDREB01, 03, 11, 14, 18 and 19) were down-regulated after 12 h. AcoDREB09 was extremely sensitive to drought stress, and its expression level quickly reached a maximum at 6 h after treatment. Except for AcoDREB06, the expression levels of the analyzed genes did not change as much in the roots as they did in the shoots. Compared to the control plants, AcoDREB03 and AcoDREB11 were rapidly down-regulated in the roots. These expression pattern changes after mannitol treatment indicate the vital role played by AcoDREB genes in response to drought conditions (Figs. 9I–9P).

Cold stress drastically affects plant growth and development and causes major crop yield losses (Cai et al., 2015). The expression levels of the DREB genes were equally affected by cold treatment in the roots and in the shoots. In particular, three genes (AcoDREB01, 03 and 18) responded rapidly to cold treatment, and their expression levels in the shoots peaked at 6 h. Two genes (AcoDREB09 and AcoDREB19) reached their maximum expression levels in the shoots after 48 h (Figs. 9Q–9X).

To analyze the effects of heat stress, the plants were subjected to 45 °C temperature. In the shoots, the majority of the analyzed genes were initially down-regulated then subsequently up-regulated. AcoDREB03 was the only gene that was up-regulated in the shoots during the first 12 h. In the roots, the expression levels of four genes (AcoDREB01, 09, 11 and 18) gradually increased and peaked at 48 h. The expression levels of two genes (AcoDREB03 and AcoDREB14) decreased rapidly after exposure to high temperature stress. Unlike the other genes, the expression of AcoDREB06 in the roots peaked at 12 h. Collectively, these results indicate the involvement of AcoDREB genes in the response to heat stress in pineapple (Figs. 9Y–9FF).