Genome-wide identification and characterization of ABA receptor PYL/RCAR gene family reveals evolution and roles in drought stress in Nicotiana tabacum

Abscisic acid (ABA) is an important phytohormone for plant growth, development and responding to stresses such as drought, salinity, and pathogen infection. Pyrabactin Resistance 1 (PYR1)/PYR1-Like (PYL)/Regulatory Component of ABA Receptor (RCAR) (hereafter referred to as PYLs) has been identified as the ABA receptors. The PYL family members have been well studied in many plants. However, the members of PYL family have not been systematically identified at genome level in cultivated tobacco (Nicotiana tabacum) and its two ancestors. In this study, the phylogenic relationships, chromosomal distribution, gene structures, conserved motifs/regions, and expression profiles of NtPYLs were analyzed. We identified 29, 11, 16 PYLs in the genomes of allotetraploid N. tabacum, and its two diploid ancestors N. tomentosiformis and N. sylvestris, respectively. The phylogenetic analysis revealed that NtPYLs can be divided into three subfamilies, and each NtPYL has one counterpart in N. sylvestris or N. tomentosiformis. Based on microarray analysis of NtPYL transcripts, four NtPYLs (from subfamily II, III), and five NtPYLs (from subfamily I) are highlighted as potential candidates for further functional characterization in N. tabacum seed development, response to ABA, and germination, and resistance to abiotic stresses, respectively. Interestingly, the expression profiles of members in the same NtPYL subfamily showed somehow similar patterns in tissues at different developmental stages and in leaves of seedlings under drought stress, suggesting particular NtPYLs might have multiple functions in both plant development and drought stress response. NtPYLs are highlighted for important functions in seed development, germination and response to ABA, and particular in drought tolerance. This work will not only shed light on the PYL family in tobacco, but also provides some valuable information for functional characterization of ABA receptors in N. tabacum.

The Static Light Scattering (SLS) and Analytical Ultracentrifugation (AUC) analysis of recombinant AtPYLs in solution revealed that AtPYR1, AtPYL1, and AtPYL2 could form homodimers, while AtPYL4, AtPYL5, AtPYL6, AtPYL8, AtPYL9 and AtPYL10 are monomers [55]. In general, these monomeric AtPYLs interact with AtPP2Cs in an ABA-independent manner, while the dimeric AtPYLs bind to AtPP2Cs in an ABA-dependent manner. The ABA-dependence of AtPYLs-AtPP2Cs interactions are determined by a conserved region which called CL2 [55]. The CL2 region forms an ABA-binding pocket with the other three highly conserved surface loops, CL1, CL3, and CL4 [56]. Therefore, the conserved CL2 region in PYLs is crucial for ABA signaling transduction in plants.
Furthermore, transcription levels of several ABA receptor genes could be regulated in different tissues and by ABA and abiotic stresses. For example, the expression profiles of Gossypium hirsutum PYLs were tissuespecific [50]. Twelve BdPYLs were identified from the monocot model plant B. distachyon genome, and expression levels of BdPYL11 are significantly down-regulated in response to ABA, NaCl, and osmotic stress [52]. Moreover, expression levels of most SlPYLs, except SlPYL1 and SlPYL8, were also down-regulated by dehydration stress in tomato leaves [54].
Cultivated tobacco (Nicotiana tabacum) is not only an important economic crop in many countries, but also widely used as a model for plant biology research. It is well known that the allotetraploid N. tabacum (2n = 48, TTSS) was originated from a hybridization event between N. tomentosiformis (2n = 24, TT) and N. sylvestris (2n = 24, SS) [57,58]. Therefore, the gene redundancy and diversity of NtPYL family might be more complex than those of AtPYL and OsPYL family in diploid Arabidopsis and rice.
Lots of studies on the ABA receptors have been carried out in many plants, mainly in Arabidopsis [8, 14-16, 30, 31, 36] and rice [37][38][39][40]. However, little is known about the ABA receptor family in cultivated tobacco [59]. In this study, we identified 29 NtPYLs in the allotetraploid N. tabacum, 11 NtomPYLs and 16 NsylPYLs in the two diploid wild tobacco species N. tomentosiformis and N. sylvestris, respectively. The phylogenic relationships, chromosomal distribution, gene structures, and the conserved motifs/regions in these PYLs were analyzed. Moreover, the expression profiles of NtPYLs in tissues at different developmental stages, and in response to drought stress were further investigated. These results provide some valuable information for further functional characterization of ABA receptors in N. tabacum.

Genome-wide analysis of PYLs in three Nicotiana species
To identify the PYL family members in the Nicotiana tomentosiformis (2n = 24, TT), N. sylvestris (2n = 24, SS), and N. tabacum (2n = 48, TTSS) genomes, the coding sequences and amino acid sequences of 14 Arabidopsis AtPYLs were applied as queries to search against the NCBI database and China Tobacco Genome Database (V2.0) in China Tobacco Gene Research Centre at Zhengzhou Tobacco Research Institute. Based on amino acid sequence similarity to AtPYLs, a total of 11, 16, and 29 PYL genes were retrieved from the genomes of two progenitor diploid species N. tomentosiformis and N. sylvestris, and the descendant tetraploid specie N. tabacum, respectively (Additional files 1, 2, 3: Tables S1, S2, S3). As showed in Table 1 Notably, the NtPYL13 and NtPYL22 are the shortest proteins (154 aa) among the NtPYLs.
The basic information of the NtomPYLs and NsylPYLs, including the gene ID, lengths of gene and ORF, MW, and pI are listed in Table 2 and Table 3 To characterize the phylogenetic relationship among PYLs from Arabidopsis, soybean, and cultivated tobacco, an unrooted neighbor-joining (NJ) tree was constructed using the MEGA software from the alignment of 29 NtPYLs in N. tabacum, 14 AtPYLs in Arabidopsis thaliana, and 23 GmPYLs in Glycine max (Fig. 1). The phylogenetic analysis of amino acid sequences from the deduced NtPYLs, PYLs in Arabidopsis and soybean revealed that the 29 NtPYLs could be grouped with their orthologous PYLs from Arabidopsis and soybean (Fig. 1). Thus, the NtPYLs    were renamed as NtPYL1 to NtPYL29 according to their sequence similarities to AtPYLs. According to the created phylogenetic tree, these PYLs could be classified to 3 subfamilies. Subfamily I include the NtPYL1-18, AtPYR1, AtPYL1 to AtPYL6, and GmPYL1 to GmPYL14. Subfamily II contains NtPYL19 to NtPYL27, AtPYL7 to AtPYL10, and GmPYL15 to GmPYL20. Subfamily III consists of NtPYL28 and NtPYL29, AtPYL11 to AtPYL13, and GmPYL21 to GmPYL23 (Fig. 1). There are 18 NtPYLs, 7 AtPYLs, and 14 GmPYLs in PYL subfamily I; 9 NtPYLs, 4 AtPYLs, and 6 GmPYLs in PYL subfamily II; 2 NtPYLs, 3 AtPYLs, and 3 GmPYLs in PYL subfamily III. Therefore, PYL subfamily I and III have the largest and minimum number of members among the Arabidopsis, soybean, and cultivated tobacco, respectively.

Chromosomal distributions of NtPYLs
The localizations of the PYLs in the chromosomes of N. tabacum were further determined. The information of a physical maps (including the length and number of each chromosome, and gene locus) in each N. tabacum chromosome were obtained from the China Tobacco Genome Database (V2.0). A simplified physical map which shows the location of NtPYLs in the N. tabacum chromosomes was drawn by the software MapGene2Chromosome (Fig. 2). In general, NtPYLs were unevenly distributed in the N. tabacum chromosomes. For example, 24 NtPYLs were distributed in the 12 chromosomes, and four NtPYLs (NtPYL2, NtPYL23, NtPYL26, and NtPYL28) were located on the chromosome 3. Each of chromosome 16 and 18 contains 3 NtPYLs (NtPYL8, NtPYL11, NtPYL12 and NtPYL4, NtPYL5, NtPYL25, respectively), while four chromosomes (chromosome 4, 6, 23, and 24) harbor 2 NtPYLs, and only 1 NtPYL were located on the each of 4 chromosomes (chromosome 7, 9, 14, and 15). However, there are 5 NtPYLs that were located on the scaffolds could not be distributed on the N. tabacum chromosomes. In addition, NtomPYLs and NsylPYLs could only be located on the scaffolds but not on the chromosomes of N. tomentosiformis and N. sylvestris, due to the lacking information of physical maps for the N. tomentosiformis and N. sylvestris.
Interestingly, according to the analysis of amino acid sequences similarity for PYLs among the N. tabacum, N. tomentosiformis, and N. sylvestris, there should be a tandem gene duplication (NtPYL11 and NtPYL12, both derive from Ntom0128520) and gene duplication (NtPYL4 and NtPYL5, both derive from Nsyl0370740) event on chromosome 16 and 18, respectively ( Fig. 2, Table 4).

Phylogenetic analysis of PYLs in three Nicotiana species
Cultivated tobacco (N. tabacum) is a tetraploid crop, and the genome of N. tabacum (TTSS) is likely derived from two different genomes of two wild diploid tobacco (T genome from N. tomentosiformis and S genome from N. sylvestris, respectively) [57,58]. To analyze the phylogenetic relationship among PYLs from N. tomentosiformis, N. sylvestris, and N.tabacum, an unrooted neighbor- Fig. 2 The chromosomal location of the ABA receptor genes (NtPYLs) on the Nicotiana tabacum chromosomes. Chromosome size is indicated by its relative length. The scale on the left is in megabases (Mb). The bars on the chromosomes indicate the positions of the ABA receptor genes. The figure was generated and modified using the MapGene2Chrom program joining tree was constructed via the MEGA software by comparing 11 NtomPYLs, 16 NsylPYLs, and 29 NtPYLs (Fig. 3, left panel). The PYLs in these tobacco species could be divided into 3 subfamilies and further classified to 13 groups, and each NtPYL has a putative orthologous gene in either N. sylvestris or N. tomentosiformis (Fig. 3, Table 4).
Notably, NtPYL4, NtPYL5 in group 2, and NtPYL22, NtPYL23 in group 10 were all derived from N. sylvestris, while NtPYL11, NtPYL12 in group 5 are originated from N. tomentosiformis. Namely, NtPYL4 and NtPYL5 in group 2 are derived from Nsyl0370740, and NtPYL11 and NtPYL12 in group 5 are derived from Ntom0128520, indicating that NtPYL4 and NtPYL5 might have the same origin from N. sylvestris, and NtPYL11 and NtPYL12 might share the same origin from N. tomentosiformis (Fig. 3, Table 4).
Notably, only NtPYL22 in the NtPYL subfamily II has one intron, and amino acid length of NtPYL22 (154 aa) is the shortest among the ABA receptors of N. tabacum (Table 1). Through comparing the length of exons and introns of members (NtPYL19 to NtPYL27) in the NtPYL subfamily II, the first exon and intron might had lost in NtPYL22 gene.

Conserved motifs and CL2 regions/loops of NtPYLs
Amino acid alignment analysis revealed that all the identified NtPYLs share a highly similar helix-grip organization with three α-helices separated by 7 β-sheets, and several conserved CL regions/loops (Fig. 4a), which have been well characterized in the PYR/PYL/RCAR ABA receptor gene family [14,15].
The conserved motifs in 29 NtPYLs were predicted via MEME and Pfam software. In total, four motifs were identified in NtPYLs. One motif (motif 4) was identified only in 17 NtPYLs from subfamily I (NtPYL1 to NtPYL18) (Fig. 4b). Most of the 29 NtPYLs share the same motifs (motif 1, 2, and 3) suggesting their conserved functions in the NtPYLs. Among the 29 NtPYLs, only 2 NtPYLs (NtPYL13 from subfamily I and NtPYL22 from subfamily II) have two motifs (motif 1 and 3). Notably, NtPYL13 and NtPYL22 are the shortest receptors (154 aa) in the NtPYL family (Table 1). Moreover, most of the members in subfamily I, except NtPYL13, have the motif 4 at the N-terminal, indicating that most members in NtPYL subfamily I might carry out special biological functions.
Previous study showed that the conserved CL2 region/ loop in PYLs is important for the interaction between Table 4 ABA receptor gene family in Nicotiana tabacum and its putative ancestors N. sylvestris and N. tomentosiformis Gene group Gene name Gene ID in database Orthologous gene Orthologous genes from N. tomentosiformis were labeled in italic NtPYL4 and NtPYL5 in group 2 share the same putative ancestor (marked with underline) from N. sylvestris. NtPYL11 and NtPYL12 in group 5 share the same putative ancestor (marked with underline) from N. tomentosiformis PYLs and PP2Cs in an ABA dependent or independent manner in Arabidopsis and soybean [15,43]. To identify CL2 region in the NtPYLs, alignment of AtPYLs, GmPYLs and NtPYLs was investigated. The conserved CL2 regions including 10 amino acid residues in the NtPYLs show a certain extent of similarity and polymorphism to those in AtPYLs, and GmPYLs (Fig. 5). The No. 3 and 4 residues in CL2 region are the amino acids that are critical for the monomeric or dimeric status, ABA dependence of PYL-PP2C interactions, and activities of AtPYLs [55]. In Arabidopsis, the combination of two key amino acid residues are VI and VV, VK and LK, VV and LV in AtPYL subfamily I, II and III, respectively. In Glycine max, the combination of two key amino acid residues are VI and VV, VK, IT and VT in GmPYL subfamily I, II and III, respectively. In Nicotiana tabacum, the combination of two key amino acid residues are VI and VV, VK and VR, LV in NtPYL subfamily I, II and III, respectively (Fig. 5). Common patterns of two key amino acid residues in the CL2 region of AtPYLs, GmPYLs and NtPYLs could be illustrated as followed: VI and VV for subfamily I, VK for subfamily II, and LV is the common pattern in subfamily III of AtPYLs and NtPYLs. The patterns of two key amino acid residues in the CL2 region in AtPYLs, GmPYLs, and NtPYLs are conserved in subfamily I and II among these three plant species, but distinguished in subfamilies III, indicating that the members in different PYL subfamilies might have similar functions in different plants.

Expression profiles of NtPYLs in tissues at different developmental stages
To investigate the expression patterns of NtPYLs in different tissues and developmental stages, the relative expression levels of NtPYLs was analyzed in N. tabacum dry seeds, germination seeds, cotyledons, leaves and roots from two, four, six, and ten true leaves stages, and flowers at squaring stage by Microarray analysis (Fig. 6, Additional file 5: Table S5).
The gene expression levels in NtPYL subfamily I have a broader range from 2.83 (NtPYL3) to 8.33 (NtPYL18) in dry seeds. When compared with the expression levels in the dry seeds, expression levels of NtPYL3, NtPYL4, and NtPYL5 were constantly higher in germination seeds, cotyledons, leaves and roots from two, four, and six true leaves stages, and leaves in the ten true leaves stage; expression levels of NtPYL10, NtPYL11, and NtPYL12 were constantly higher in cotyledons, leaves, and roots in two true leaves stage; expression levels of NtPYL15 and NtPYL16 were higher from germination seeds to roots in the two true leaves stage; expression levels of NtPYL6 and NtPYL18 were constantly lower from four true leaves stage to flowering stage. Notably, the NtPYL7 and NtPYL17 had lower expression levels in leaves comparison with roots in the four, six, and ten true leaves stage.
In NtPYL subfamily II, the expression levels of most genes (8 of 9 genes) ranged from 5 to 8, except the lower expression level of NtPYL23 (3.3) in dry seeds. Compared with the expression levels in the dry seeds, expression level of NtPYL23 showed constantly higher in germination seeds till flowers, and expression level of NtPYL26 and NtPYL27 have constantly higher in cotyledons till leaves in ten true leaves stage. However, NtPYL21 and NtPYL22 showed lower expression levels in leaves of four, six, and ten true leaves stage and flowers than those in dry seeds. Notably, expression levels of NtPYL21 and NtPYL22 are lower in leaves compared with roots at four, six, and ten true leaf stages.
For the two members of NtPYL family III, NtPYL29 (5.13) showed higher expression level than that of NtPYL28 (3.84) in dry seeds. Interestingly, expression levels of both NtPYL28 and NtPYL29 are constantly lower in all the developmental stages after dry seeds stage compared with those in the dry seeds stage (Fig. 6, Additional file 5: Table S5).

Expression profiles of NtPYLs in response to drought stress
To understand the possible function of NtPYLs in plant response to drought stress, we analyzed the expression profiles of NtPYLs in the tobacco seedlings after drought treatment for indicated time. The expression levels and patterns of all NtPYLs were detected by the Microarray. Notably, three pairs of gene, NtPYL4 and NtPYL5, NtPYL11 and NtPYL12, and NtPYL26 and NtPYL27, have identical expression levels and patterns (Fig. 7, Additional file 6: Table S6). The NtPYL4 and 5 are derived from Nsy10370740, while NtPYL11 and 12 are derived from Ntom0128520 (Fig. 3, Table 4). The amino acid sequence similarity between NtPYL26 and 27 is 98.87% (Additional file 4: Table S4). Therefore, the microarray experiment could not distinguish the expression profiles for each of three gene pairs.
In addition, the expression value of each NtPYL in CK was assigned as 1, and the expression ratio for the treatment at indicated time/CK was calculated for the expression pattern of each NtPYLs (Additional file 6: Table  S6). The ratio between treatment and CK that is more than 1.2 or less than 0.8 was recognized as up-or downregulation, respectively. For the NtPYLs in subfamily I, expression levels of 4 NtPYLs (NtPYL6, NtPYL15, NtPYL16 and NtPYL18) are down-regulated after dehydration treatment for 2 h, expression levels of 7 NtPYLs (NtPYL7, NtPYL8, NtPYL9, NtPYL10, NtPYL11, NtPYL12 and NtPYL17) are up-regulated after dehydration treatment for 0.5 h and down-regulated thereafter, and expression levels of 7 NtPYLs (NtPYL1 to NtPYL5, NtPYL13 and NtPYL14) were changed slightly. For the NtPYLs in subfamily II, expression levels of 2 NtPYLs (NtPYL24, 25) are up-regulated after dehydration (See figure on previous page.) Fig. 4 Alignment and conserved motifs of NtPYLs. Amino acid sequence alignment of the 29 NtPYLs and AtPYL2 was performance by ClustalW. a Secondary structural elements are indicated above the primary sequence. Helices and sheets/strands are shown as black helices and arrows, respectively. The four conserved ABA receptor region CL1-CL4 are indicated with red lines. The conserved motifs analysis of the NtPYLs based on their phylogenetic relationship were identified using MEME software. b In left panel, the members of each subfamily are indicated with the same color and different NtPYL subfamilies are represented by the Roman numeral I-III in the phylogenetic tree. In right panel, grey lines represent non-conserved sequences, and colored boxes numbered at the bottom indicate different motifs. The length of motifs in each NtPYL protein is shown proportionally treatment for 2 h. For the subfamily III, the expression level of NtPYL29 is slightly up-regulated after dehydration treatment for 8 h (Fig. 7, Additional file 6: Table  S6). Notably, the expression levels of most NtPYLs (11 NtPYLs in 18 NtPYLs) in subfamily I are decreased after dehydration treatment for 2 h, which is consistent with the previous study in tomato [54].
To further confirm the microarray data, gene expression of NtPYLs were performed by quantitative real-time PCR. The specific primers of NtPYLs were used for qRT-PCR  Table S7). Consisted with microarray data, gene expressions of NtPYL6, NtPYL16 and NtPYL18 were down-regulated in response to dehydration treatment, and expression levels of NtPYL9, NtPYL11 and NtPYL17 were up-regulated for 0.5 h after dehydration treatment. While gene expressions of NtPYL1, NtPYL3, NtPYL4 and NtPYL14 were not significantly changed under dehydration treatment (Fig. 8) . In addition, transcript levels of NtPYL24 and NtPYL25 were up-regulated after dehydration treatment, which is consisted with microarray data (Fig. 8).
According to the coding and amino acid sequences of 14 AtPYLs, 11 NtomPYLs, 16 NsylPYLs, and 29 NtPYLs were retrieved from the genomes of the tetraploid specie N. tabacum and its two progenitor diploid species N. tomentosiformis, and N. sylvestris, respectively. NtPYLs, NtomPYLs, and NsylPYLs and their putative proteins exhibited similar physical properties. In addition, amino acid sequence alignment analysis Fig. 6 The expression profile of 29 NtPYLs in tissues at different developmental stages. The relative transcript abundances of 29 NtPYLs were examined via microarray and visualized as a heatmap. The gene expression profiles of NtPYLs in 12 different samples, including dry seeds, germination seeds, cotyledons, leaves from two-true leaf stage (labeled as two true leaf_leaf), roots from two-true leaf stage (two true leaf_root), leaves from fourtrue leaf stage (four true leaf_leaf), roots from four-true leaf stage (four true leaf_root), leaves from six-true leaf stage (six true leaf_leaf), roots from sixtrue leaf stage (six true leaf_root), leaves from 10-true leaf stage (ten ture leaf_leaf), roots from ten-true leaf stage ten ture leaf_root), and flowers at squaring stage (squaring stage_flower). The X axis is the samples in tissues at different developmental stages. The color scale represents Log2 expression values showed that all the 29 NtPYLs share highly similar structure characterized by several α-helices, β-sheets and conserved CL loops, similar with the AtPYL2, a functional member of the ABA receptor family in Arabidopsis [62]. Therefore, these NtomPYLs, Nsyl-PYLs, and NtPYLs are the ABA receptors in N. tomentosiformis, N. sylvestris, and N. tabacum, respectively. In cotton, PYL families had been identified recently at genome level in two ancestral diploid species Gossypium raimodii and G. arboretum, and two tetraploid species G. barbadense and G. hirsutum derived from G. raimodii and G. arboretum, respectively. Analysis of the physical properties of the 20 GrPYLs, 21 GaPYLs, 39 GrPYLs, and 40 GaPYLs revealed that the Gossypium PYL families share similar ORF and amino acid lengths, MW and pI [50]. Compared with the published plant PYL families [14,15,38,41,48,50], we found that the Nicotiana PYLs had similar physical properties regarding to the amino acid lengths, ORF and MW. Notably, the number of PYL members (29) in N. tabacum is larger than those in the tetraploid soybean (23) [43] and other diploid plant species, such as Arabidopsis (14) [14][15][16], and tomato (15) [41], but lesser than that in the tetraploid cotton (39 or 40) [50], indicating that tetraploid plant species that harbor large PYL families might have more complex ABA responses in plant development and respond to environmental stress.
In Arabidopsis and soybean, PYLs could be classified into 3 subfamilies, and each subfamily might have diverse function [14,15,43]. The phylogenetic analysis of amino acid sequences of NtPYLs, AtPYLs and GmPYLs showed that NtPYLs could also be grouped into 3 subfamilies. Further analysis also identified 3 subfamilies among three Nicotiana species (N. tomentosiformis, N. sylvestris, and N. tabacum). Amino acid sequence alignment of Nicotiana PYLs suggested that these PYLs could be further divided into 13 groups ( Table 4). The phylogenetic relationships of Nicotiana PYLs revealed that each of the 29 NtPYLs had an orthologous ancestor in either N. tomentosiformis or N. sylvestris. Notably, 16 Nsyl-PYLs and 11 NtomPYLs have been identified in N. sylvestris and N. tomentosiformis, respectively. These results suggested that both S-genome from N. sylvestris (SS) and T-genome from N. tomentosiformis (TT) had contributed the PYLs for the genome of N. tabacum (TTSS), and S-genome might contribute more PYLs than that of T-genome during evolution [57,58].
It was showed that NtPYL4 and NtPYL5 are located in the different positions of chromosome 18 in the N. tabacum genome, while NtPYL11 and NtPYL12 are  Gene structures of Nicotiana PYLs could also provide useful information to understand the evolution relationship of PYL family from N. sylvestris and N. tomentosiformis to N. tabacum. For example, since NtPYL13, NtPYL22 is the shortest Nicotiana PYLs (154 aa) identified in this study (Fig. 3, Fig. 4 a), alignment analysis of the gene structures and conserved motifs of NtPYL13, NtPYL22, and their putative ancestors revealed that both NtPYL13 and NtPYL22 are lacking the N-terminal motifs when compared with their putative ancestor Ntom0177100 and Nsyl0129950, respectively (Fig. 4 b). These results indicated that NtPYL13 had lost the 5′-region of Ntom0177100, and NtPYL22 had lost the first exon and intron of Nsyl0129950 during the evolution.
The CL2 loops/regions, particularly the combination of two key amino acid residues in the CL2 region, in AtPYLs determine the ABA-dependence of PYL-PP2C interactions in Arabidopsis [55,56]. Recent studies on the mutation and degradation of PYLs for PYL-PP2C interactions, together with the downstream substrates of SnRKs, suggest the potential to engineer the core PYLs-PP2Cs-SnRKs components of ABA signaling [30,32,[63][64][65]. For the combination of two key amino acid residues in the CL2 region, NtPYLs share the same pattern with those in the AtPYLs, GmPYLs: VI and VV for PYL subfamily I, and VK for PYL subfamily II. AtPYL13, NtPYL28, and NtPYL29 of the PYL subfamily III share the pattern of LV. The conserved combination of two key amino acid residues in the CL2 region in N. tabacum, Arabidopsis and soybean, suggest that the CL2 loop/region play vital roles in PYL-PP2C interactions in plants. Whether N. tabacum have the similar ABA dependence of PYL-PP2C interactions as those in Arabidopsis deserves further study using yeast two-hybrid and bimolecular fluorescence complementation (BiFC) assays.
Expression profiles analysis in many plants, such as Arabidopsis [5,35,36], rice [37,38,40], tomato [41,54], maize [46], cotton [50], and rubber tree [48], revealed that the majority of PYLs are expressed in various tissues including seeds, root, leaves, flowers and fruits. For example, most PYLs were expressed in all tissues of rice, and OsPYL7/8 were highly expressed in embryos, OsPYL3 and OsPYL5 were primarily expressed in leaves, while OsPYL1 was predominantly expressed in roots [38]. In solanaceous tomato, most of the PYL genes (including two members of subfamily I, five members of subfamily II, and two members of subfamily III) were highly expressed in roots, while one member (Sl1g095700) from PYL subfamily III showed the highest expression levels than those of other PYLs in leaves. Notably, one gene (Sl6g061180) from PYL subfamily I and most of the members in PYL subfamily II and III exhibited high expression levels during tomato fruit development. Interestingly, transcripts of many fruit-expressed SlPYLs gradually increased to the highest level and declined afterwards. Sun et al. also reported that transcripts of SlPYL1, SlPYL2 (two tomato orthologue of AtPYL7 and AtPYL9), SlPYL3 (a tomato orthologue of AtPYL8 and AtPYL10), and SlPYL6 (a tomato orthologue of AtPYL4) within the SlPYL family fluctuated during fruit development and ripening in tomato [54]. Additionally, expression profile analysis of PYLs in a tetraploid cotton (G. hirsutum) demonstrated that 22 GhPYLs were preferentially expressed in flowers, 10 GhPYLs were dominantly expressed in roots, and 3 GhPYLs were highly expressed in the fiber [50]. As a tetraploid solanaceous specie, N. tabacum might have similar tissue-and developmental stage-specific expression of PYLs as those reported in tomato and cotton.
In this study, expression profiles of NtPYLs in different tissues and developmental stages were analyzed by chip-Microarray. Transcripts of 29 NtPYLs could be detected in all the tissue/developmental stage checked (Fig. 6, Additional file 5: Table S5). Notably, NtPYL28 and NtPYL29 from subfamily III showed a distinct expression pattern during the development: transcripts of NtPYL28 and NtPYL29 reached the highest level in dry seeds, reduced dramatically in germination seeds, and remain almost constant low level afterwards. The highly accumulation of NtPYL28 and NtPYL29 transcripts particularly in dry seeds, suggest that NtPYL28 and NtPYL29 might play important roles in N. tabacum seed dormancy. Kim et al. reported that overexpressing OsPYL/RCAR5 (a rice orthologue of AtPYL8) resulted in hypersensitive to ABA during seed germination and early seedling growth [37]. Overexpressing OsPYL3 (a rice orthologue of AtPYL8) and OsPYL9 (a rice orthologue of AtPYL3) confer ABA hypersensitivity during seed germination in rice [38]. These studies in rice indicated some OsPYLs (OsPYL3, 5, and 9) are involved in seed germination and early seedling growth in response to ABA treatment. Recently, AtPYL8 and AtPYL9 have been shown to play important roles in regulating lateral root growth in the presence of ABA [5,36]. Given NtPYL19 and NtPYL20 are cluster closely with AtPYL7, AtPYL9, AtPYL8 and AtPYL10 in phylogenetic analysis (Fig. 1), together with the previous studies on functions of the orthologue of AtPYL8 and AtPYL10 in tomato fruit development, and unique expression patterns of NtPYL28 and NtPYL29 in dry seeds during developmental process, the functions of NtPYL19, 20, 28, 29 in seed development, germination, response to ABA deserve further study.
Since the phytohormne ABA play crucial roles in plant responses to drought, we analyzed the expression levels of NtPYLs in the whole seedling after dehydration treatment using chip-Microarray (Fig. 7, Additional file 6: Table S6). In general, the transcripts of members in NtPYL subfamily II and III remain constant in the control seedlings (CK, without drought stress) and during the drought treatment, indicating that NtPYL19-27, NtPYL28, and NtPYL29 might not involve in the N. tabacum seedling responses to drought stress. Interestingly, for members in NtPYL subfamily I, the transcriptional responses to drought treatment could be divided into three types: (1) NtPYL1-5, Overexpression OsPYL/RCAR5 (a rice orthologue of AtPYL8), OsPYL3 (a rice orthologue of AtPYL8), and OsPYL9 (a rice orthologue of AtPYL3) conferred improved drought stress tolerance in rice [37,38,40]. Moreover, in leaves of tomato seedlings subjected to dehydration, transcript accumulation of 6 SlPYLs (SlPYL2-7) were deduced, and expression levels of 3 SlPYLs (SlPYL4, 6, and 7) were recovered after re-watering [54].
The unexpected results in Arabidopsis showed that expression levels of AtPYLs were down-regulated by stress [12]. Overexpression of AtPYL5 and AtPYL9 resulted in enhanced ABA responses and drought resistance in Arabidopsis through PYL-mediated inhibition of clade A PP2Cs [16,34,35]. Importantly, Gonzalez-Guzman et al. revealed that overexpression of tomato monomeric-type, but not dimeric ABA receptors in Arabidopsis confers enhanced resistance to drought stress [41]. In this study, NtPYL6-12 in NtPYL subfamily I were clustered closely with AtPYL4 and AtPYL5 (Fig. 1), which are monomeric-type ABA receptors [55]. Together with the NtPYLs that showed down-regulated expression pattern in NtPYL subfamily I, NtPYL6, 7, 10, 11, 12 will be the key candidate ABA receptors for functional identification in N. tabacum response to drought, salt, and osmotic stresses.
Notably, a comprehensive comparison comparing transcriptional profiles of the core ABA signaling components under osmotic/dehydration stress or ABA treatment between roots and leaves of maize (Zea mays) seedlings grew in hydroponic culture revealed that, after treating roots with ABA, the expression of ZmPYLs homologous to monomeric-type AtPYLs were reduced, whereas those of ZmPYLs homologous to dimeric-type AtPYLs were increased in maize primary roots. Surprisingly, the opposite pattern was observed in the leaves of the same experiments [46]. This interesting organ-specific expression patterns for ABA receptor genes between roots and leaves of maize in response to ABA suggest that there might be a contrast transcriptional patterns for monomeric-and dimeric-type NtPYLs in roots and leaves under abiotic stresses and ABA treatment, which deserves further study in the future study.
Interestingly, the expression patterns of many members in each NtPYL subfamily are similar in different tissues/ developmental stages and in seedlings under drought stress, suggesting these NtPYLs could be regulated by the developmental signals and dehydration treatment as well. Characterization of these NtPYLs is required for understanding their functions during developmental process and in response to stresses.
In the current study, our study identified 29 putative NtPYLs in the allotetraploid cultivated tobacco (N. tabacum), 11 NtomPYLs and 16 NsylPYLs in the two diploid ancestral tobacco species, N. tomentosiformis and N. sylvestris, respectively. We further investigated the physical properties, phylogenetic relationship, protein motifs, and gene structures of Nicotiana PYLs. The Nicotiana PYLs could also be divided to 3 subfamilies, in consistent with the results from studies in other plants. Phylogenetic, gene structure, and protein motif analysis revealed NtPYLs were originated from NtomPYLs and NsylPYLs, NtPYL22 had lost the first exon and intron compared with its origin in N. sylvestris0129950, and NtPYL13 was derived from N-terminal truncation of N. tomentosiformis0177100 during evolution. Moreover, analysis of NtPYLs expression profiles via chip-Microarray assay indicated that each NtPYL subfamily might have diverse functions in different tissues/developmental stages and in response to abiotic stresses. Furthermore, four NtPYLs (NtPYL 19, 20, 28, and 29) in NtPYL subfamily II and III were suggested to play important roles in seed development, germination, and response to ABA. Finally, five NtPYLs (NtPYL6, 7,10,11,12) in NtPYL subfamily I were highlighted as potential candidates for further characterizing their functions in N. tabacum resistance to abiotic stresses. Taken together, the results from genome-wide identification of Niacotiana PYLs will provide some insights on understanding the roles of PYLs in ABA signaling and in response to abiotic stresses in tetraploid plants, which will facilitate the improvement of crop resistance to drought stress.

Conclusions
ABA receptors (PYLs) play central roles in ABA signaling and plant response to many environmental stresses. Although lots of PYL genes and family have been identified in many plant species, the information of Nicotiana PYL family is still missing. Here we conducted a genome-wide identification and expression analysis of the PYL family in N. tabacum. A total of 29, 11, 16 Nicotiana PYL genes were identified in the genome of N. tabacum and its two ancestors N. tomentosiformis and N. sylvestris, respectively. Phylogenetic and gene structure analysis revealed that Nicotiana PYL could be divided into 3 subfamilies and 13 groups. Furthermore, each NtPYL might have a putative orthologous gene in either N. sylvestris or N. tomentosiformis, in consistent with the evolutionary origin of N. tabacum. The microarray-based analysis of NtPYLs expression profiles in tissues at different developmental stages, and in response to drought stress revealed that members in different NtPYL subfamily might paly specific roles in N. tabacum growth, developmental, and drought stress responses. Interestingly, the expression profiles of members in the same NtPYL subfamily showed somehow similar patterns in tissues at different developmental stages and in leaves of seedlings under drought stress, suggesting particular NtPYLs might have multiple functions in both plant development and drought stress response. Importantly, four NtPYLs (NtPYL 19, NtPYL20, NtPYL28, and NtPYL29) are highlighted for the potential functions in seed development, germination and response to ABA. Moreover, five NtPYLs (NtPYL6, NtPYL7, NtPYL10, NtPYL11, NtPYL12) might play important roles in response to abiotic stresses, particularly in drought. Taken together, these results will facilitate further functional characterization of NtPYLs in plant development and in response to abiotic and biotic stresses in tetraploid plants.

Plant materials and growth conditions
The Nicotiana tabacum L. cv. Honghuadajingyuan seeds were obtained from Yunnan Academy of Tobacco Agricultural Sciences (Yunnan, China) [66]. Surfacesterilized seeds were directly sowed into the soil in pots. The tobacco young seedlings were grown in the plant growth chamber with a 16-h-light/8-h-dark photoperiod under continuous white light (∼75 mol m − 2 s − 1 ) at 28°C-day/ 23°C-night. All plants were kept well-watered after sowing.
For expression profiling of ABA receptor genes in response to drought stress, the plants were grown for 7-8 weeks with 6-7 leaves. The plants were moved out from the pots carefully without disturbing the root, and the surface soil was washed out softly. Then the plants were put on the bench for air drying which termed as drought stress treatment. The whole seedlings were collected at the indicated time after treatment, and immediately frozen in liquid nitrogen for RNA extraction for microarray assay. Five biological replicates were used for sample harvesting at each indicated time of the treatment.
For expression profiling of ABA receptor genes in different developmental stages and tissues in tobacco, tobacco plants were kept in the growing condition mentioned above. Samples were harvested from 12 tissues at different developmental stages, including dry seed, germination seeds, cotyledon, leaves and roots from two, four, six, and ten true-leaf stage, respectively, and flowers at squaring stage. Samples were immediately frozen in liquid nitrogen for RNA extraction for microarray assay. Three to five biological replicates were used for sample harvesting.
Analyses of phylogenetic, conserved motif, isoelectric point prediction, gene structure, and chromosome localization The Nicotiana PYL gene sequences were retrieved from NCBI (https://www.ncbi.nlm.nih.gov/) and China tobacco genome database V2.0 (data not published). The genomic DNA, open reading frame, and deduced protein sequences of Nicotiana PYL family are provided in the (Additional files 1, 2, 3: Tables S1, S2, S3), and had been submitted to NCBI database (currently waiting for assigning the accession numbers).
The sequences of GmPYLs and AtPYLs were retrieved from the NCBI GeneBank database. The sequences of NtPYLs, AtPYLs, and GmPYLs were aligned using Clus-talW [67], and an unrooted phylogenetic tree was generated using MEGA 7.0 software (http://www.megasoftware. net) by the neighbor-joining method with 1000 replicates of bootstrap analysis. For analyzing secondary structure of NtPYLs, the above alignment results were further treated by ESPript 3.0 software with default parameter settings (http://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi).

Transcriptomic microarray analysis
Total RNA was extracted with SuperPure Plantpoly RNA Kit (GeneAnswer, China). All RNA samples were treated with RNase-free DNase I (GeneAnswer) and analyzed for integrity on a Bioanalyzer 2100 (Agilent technologies, USA). About 33.3 ng total RNA were used for amplification with WT Amplification Kit (Affymetrix, Thermo Fisher Scientific, USA). 5.5 μg of the amplified product were fragmented by uracil-DNA glycosylase and apurinic/ apyrimidinic endonuclease 1 (Affymetrix, Thermo Fisher Scientific, USA).
The fragmented cDNA was labeled by terminal deoxynucleotidyltransferase using the DNA Labeling Reagent (Affymetrix, Thermo Fisher Scientific, USA) that was covalently linked to biotin. The resulting labeled cDNAs (5.2 μg) were dissolved in 160 μl of hybridization mix solution, then denatured at 99°C for 5 min. The mixed hybridization buffer was loaded into a microarray, and then the both septa were covered by round labels to prevent leaks and evaporation.
An Affymetrix custom Tobacco Genome Array with feature Size 5 μm was used. Eighty thousand six hundred fifty-two tobacco genes were covered within this array. Tobacco L25, EF1-alpha, Ntubc2, PP2A genes were used as housekeeping genes. RMA method provided by the R package, affy package, was used to conduct background correction, normalization, probe-specific background correction, probe summarization and to convert probe level data to expression values.
The hybridizations were performed in a hybridization oven (Affymetrix, Thermo Fisher Scientific, USA) at 45°C for 16 h. After hybridization, microarrays were washed by Fluidics Station 450 with wash buffer A and B (Affymetrix, Thermo Fisher Scientific, USA). Three biological replicates were used in the Microarrays assay. The expression levels of members in NtPYL family in several tissues at different developmental stages and in response to drought stress were documented in Additional file 5: Table S5 and Additional file 6: Table S6, respectively.

qRT-PCR validation of chip-microarray
For RT-qPCR validation of expression pattern of ABA receptor genes in response to drought stress, the plants were grown for 7-8 weeks with 6-7 leaves. The plants were moved out from the pots carefully without disturbing the root, and the surface soil was washed out softly. Then the plants were put on the bench for air drying which termed as drought stress treatment. The whole shoots were collected at the indicated time after treatment, and immediately frozen in liquid nitrogen for RNA extraction. Three to five biological replicates were used for sample harvesting at each indicated time of the treatment.
RNA was extracted from three to five biological replicates of the whole shoots collected at the indicated time after drought treatment using the Qiagen RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) following the manufacturer instructions.
2 μg of total RNA in a 20 μl reaction was converted to cDNA with a SuperScript III Reverse Transcriptase (Invitrogen, USA) by manufacturer instructions on a Eppendorf Mastercycler thermocycler (Eppendorf AG, Germany) with the following conditions: 25°C for 5 min, 50°C for 60 min, 70°C for 15 min, followed by a hold at 4°C until use in RT-qPCR reaction. 60 μl deionized water was added into 20 μl cDNA, and 1 μl of diluted cDNA mixture was used as the input for qPCR reaction. qPCR reactions were made with a SuperReal PreMix Plus SYBR Green Kit (TIANGEN Biotech, China) following manufacturer instructions in a 20 μl volume. The specific primers of NtPYLs were used for qRT-PCR validation (Additional file 7: Table S7).
qPCR was done on an Applied Biosystems™ QuantStu-dio™ 6 Flex Real-Time PCR System (ThemoFisher Scientific, USA) with the following cycling conditions: 95°C for 15 min, followed by 40 cycles of 95°C for 10 s, 60°C for 20 s, and 72°C for 32 s. Melt curve conditions were 95°C for 15 s, 60°C for 1 min, 95°C for 15 s. All samples had only one melt temperature peak. Fold change between experimental samples (with drought treatment) and control samples (without drought treatment) was calculated by the 2 -ΔΔCT method using 26S as a reference gene. CT values represent the average of three technical replicates.

Statistical analysis
The presented values are the means ± SE of three individual experiments with three replicated measurements. An analysis of variance (ANOVA) was used to compare significant differences based on significance levels of P < 0.05 and P < 0.01.