Characterization of Dof Transcription Factors and Their Responses to Osmotic Stress in Poplar (Populus trichocarpa)

The DNA-binding One Zinc Finger (Dof) genes are ubiquitous in many plant species and are especial transcription regulators that participate in plant growth, development and various procedures, including biotic and abiotic stress reactions. In this study, we identified 41 PtrDof members from Populus trichocarpa genomes and classified them into four groups. The conserved motifs and gene structures of some PtrDof genes belonging to the same subgroup were almost the same. The 41 PtrDof genes were dispersed on 18 of the 19 Populus chromosomes. Many key stress- or phytohormone-related cis-elements were discovered in the PtrDof gene promoter regions. Consequently, we undertook expression profiling of the PtrDof genes in leaves and roots in response to osmotic stress and abscisic acid. A total of seven genes (PtrDof14, 16, 25, 27, 28, 37 and 39) in the Populus Dof gene family were consistently upregulated at point in all time in the leaves and roots under osmotic and abscisic acid (ABA) stress. We observed that 12 PtrDof genes could be targeted by 15 miRNAs. Moreover, we mapped the cleavage site in PtrDof30 using the 5’RLM-RACE. The results showed that PtrDofs may have a role in resistance to abiotic stress in Populus trichocarpa.


Introduction
Respond to specific signals by gene expression to resist damage and to survive in complicated environments, including specific metabolic and physiological pathways. Abiotic stresses, such as drought treatment and irregular temperature, strongly impact the development and growth of plants and have adverse effects on production and quality. In general, regulation of particular gene transcription has a critical impact on many biological and evolutionary processes in plants, for instance, stress and hormone responses. It is evident that transcription factors (TFs) are essential components that regulate transcriptional rates of their target genes by binding to cis-regulatory elements in the promoters.
The DNA-binding One Zinc Finger (Dof) transcription factor family contains a zinc finger domain, and plays critical roles as plant-specific transcriptional regulators in vital a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 processes and functions in higher plants, such as stress response, photosynthetic carbon assimilation, dormancy and seed germination. Dof transcription factors typically contains 200-400 amino acids, among them, there was 52 amino acid with high level of conservation at the N-terminal, which has been considered as a DNA-binding domain. This domain is characterized by the structure of a Cys2/Cys2(C2/C2) zinc finger binding specifically to the cis-regulatory element comprising the common core sequence (AT)/AAAG [1][2]. The DNA-binding domain, which binds DNA and interacts with other proteins, is a vital functional domain [3]. The C-terminus of the Dof proteins contains a transcriptional regulation domain with various functions, including interaction with diverse regulatory proteins and the activation of gene expression [4]. The N-and C-terminal regions of the Dof protein may interact with diverse regulatory proteins or intercept signals to mediate the activation or repression of the target genes [5].
The Dof genes are ubiquitous in many plant species and are plant-specific transcription regulators that are involved in various abiotic processes. The DOF protein (ZmDOF1) first identified in maize plays a role in light-regulated gene expression, but no Dof genes have been isolated from other eukaryotes, such as yeast or humans [5]. In Arabidopsis, some of the well characterized Dof genes were shown to be involved in many plant biological processes. For example the DAG1 genes had the active involvement in seed germination [6]; the CDF1, CDF2, CDF3 and CDF5 genes are associated with photoperiodic regulation of flowering [7][8]; and HPPBF3, COG1 and OBP3, participate in the regulation of phytochrome signaling [9][10][11]. In rice, OsDof3 regulates the expression of gibberellins, OsDof12 and OsDof23, regulate flowing time and seed expression, respectively, and OsDof24 and OsDof25 are involved in carbon and nitrogen metabolism [12][13]. In tobacco, the Sar8.2b gene can be activated by the Dof transcription factor, which is related in systemic acquired resistance [14]. In sorghum, SbDof genes are associated with the responsiveness to light, hormones and endosperm-specific genes [15]. In maize, Dof1 plays a regulatory role in controlling gene related to carbohydrate metabolism [2]. In wheat, TaDof1 is related to carbon metabolism by increasing the regulation of the C4 pathway [16]. WPBF of wheat is involved in growth and development processes [17]. In addition, Dof genes are involved in physical interactions with other TFs such as bZIP, MYB, and WKRY, implying they have been implicated in the regulation of plant physiological processes.
To date, few studies are concerned with Dof gene family in P. trichocarpa, compared with the comprehensive researches of these genes in other plant species. Moreover, previous research rarely mentioned the responses of Populus Dof genes to drought stress. Therefore, there dose require a comprehensive analysis of the Dof gene family in P. trichocarpa. In this study, 41 Dof genes were systematically studied, along with their gene structures and promoter cis-elements. Their expression profiles in the leaves and roots of Populus under drought stress were examined using heatmap data and gene expression analysis with quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) analyses under osmotic and abscisic acid (ABA) treatments. Our study is the first time to provide insight about the role of PtrDof genes in stress response in Populus.
Populus trichocarpa is a valuable forest resource used widely to produce various paperbased and timber products, and has genuine commercial and ecological value. However, recent drought induced forest mortality became more and more serious, especially for Poplar. Thus, identifying PtrDof genes will provide a novel insight for drought stress resistance in Populus.

Materials and Methods
Identification and characteristics of the Dof gene family Data for the P. trichocarpa genome, including the protein database and the genomic and cDNA libraries, were obtained from the Phytozome v9.1 (http://www.phytozome.net/search. php) and NCBI (http://www.ncbi.nlm.nih.gov/) databases. We searched for the Dof genes of P. trichocarpa using two verified methods. The first method was using the Protein family (Pfam 27.0) database (http://pfam.sanger.ac.uk/), the second method was using the Hidden Markov Model (HMM) profile of the Dof gene family (protein family ID: PF02701). We verified all of the located sequences by additional manual analysis to ensure a zf-Dof domain does exist with the SMART (http://smart.emblheidelberg.de/) database [18]. All available genes were contrasted with the Dof gene family in PlnTFDB v3.0 (http://plntfdb.bio.uni-potsdam.de/v3.0/) to avoid missing genes [19]. Any additional genes were analyzed further. The ExPasy (http://web. expasy.org/protparam/), an online program was used in calculating the molecular weight and isoelectric point (pI) [20]. The subcellular localization of the Dof genes was predicted with WoLF PSORT (http://wolfpsort.org/) [21].

Chromosomal location
The chromosomal locations of genes were mapped by the PopGenIE v3 database (http://www. popgenie.org/) [26]. A physical map was constructed with Adobe Illustrator CS5 (Adobe Systems Incorporated). Genes are separated by five or fewer gene loci with a 100 kb distance were defined as tandem duplicates [27].

Gene Ontology (GO) annotation
Blast2GO v3.0 was used to analyze the functional classfication of Dof sequences and obtain the details of the annotation results [29]. Genes are described according to three classifications of GO categories: biological processes, molecular functions and cellular components.

ExHeatMap analysis
The PopGenIE v3 database (http://www.popgenie.org/) was used to obtain the expression profiles of Populus Dof genes in leaves and roots under drought stress. Detailed descriptions of Dof gene expression in response to drought stress were downloaded from the exPlot tool at the PopGenIE v3 database. Cluster and java were used to analyze these data.

Plant material and treatment
P. trichocarpa (genotype Nisqually-1) was donated by professor Liquan Jiang of North Carolina State University. P. trichocarpa was clonally propagated by culture in half-strength Murashige and Skoog medium (1/2 MS, pH 5.8) under long-day conditions (16 h light / 8 h dark) at 25˚C. Before determining the final concentration of mannitol, we performed the preliminary experiment to evaluate the effect of different concentrations of mannitol on plants growth in vitro. The results showed that 150 mM mannitol could reduce but not abolish growth of plants in vitro. For osmotic stress, plantlets were exposed to 150 mM mannitol. Plants cannot take in water normally due to the high osmotic pressure of mannitol. Thus, the mannitol simulated osmotic stress by increasing osmotic shock. For ABA stress conditions, plantlets were treated with 200 μM ABA. Mannitol was added to 1/2 MS medium at the stated concentrations and then autoclaved (115˚C, 15 min). ABA was filter-sterilized and then mixed into 1/2 MS medium at the stated concentrations. The 150 mM mannitol and 200 μM ABA treatments lasted for 0, 3, 6, 12, or 24 h or 7 d. Young leaves and roots were collected as samples at all time points. Each stress treatment performed with three biological replicates. All samples were quickly frozen in liquid nitrogen and stored at −80˚C until using. Untreated plants were used as controls.

RNA isolation and qRT-PCR verification
Total RNA was extracted using the cetyltrime thylammonium bromide method from roots and leaves [30]. The cDNA synthesized using the TransScript 1 One-Step gDNA Removal and cDNA Synthesis SuperMix. The specific sequences of each gene were based on results from multiple sequence alignment (http://multalin.toulouse.inra.fr/multalin/multalin.html). The specific primers for Dof genes were designed by Primer Premier 5 according to its CDS with following parameters, melting temperatures of 58-62˚C, primer lengths of 18-22 bp and product lengths of 190-210 bp. The primer details are listed in S1 Table. QRT-PCR was carried out using a TransStart 1 Top Green qPCR SuperMix (TransGen Biotech, Beijing, China) to determine the transcript levels under stress. Reactions were performed in 20 μL volume containing 10 μL 2×TransStart 1 Top Green qPCR SuperMix, l μL cDNA template, 7 μL ddH 2 O, and 2 μL each primer-specific. Three experimental replicates were used for each sample to maintain accuracy. The P. trichocarpa Actin1 gene (GenBank ID: XM_002298674) was used as the reference gene [31]. Other specifications, including PCR conditions and relative gene expression calculations we based on previous study [32]. All reactions were carried out under the following PCR conditions: an initial denaturation step of 95˚C for 1 min, a three-step thermal cycling profile of denaturation at 95˚C for 5s, primer annealing at 55˚C for 30 s, and extension at 72˚C for 30 s. Then, an additional step of 80˚C for 1 s was performed to remove primer dimers, followed by plate reading. We used the relative quantification method (2 −ΔΔCT ) to evaluate relative gene expression between replicates.

Statistical analysis
All the statistical analysis used one-way ANOVA to determine significance. Tukey's test was analyzed to compare the difference. Significance was defined as Ã P < 0.05, ÃÃ P < 0.01. miRNA target analysis and target validation by RLM-RACE PMRD (http://bioinformatics.cau.edu.cn/PMRD) was used to download the mature P. trichocarpa miRNA sequences. The Plant Small RNA Target Analysis Server (psRNA Target: http:// plantgrn.noble.org/psRNATarget) was used to identify the miRNA target genes in the PtrDof families with default parameters. FirstChoice RLM-RACE Kit (Invitrogen, Thermo Fisher Scientific) was used to perform RLM-RACE and illustrate the predicted targets following the methods researched by Song et al [33]. The gene-specific primers for RLM-RACE are presented in S2 Table. The RLM-RACE products were ligated into the pMD18-T vector (TaKaRa), and sequenced.

Identification of Dof genes in Populus
We used the HMM profile of the Pfam Dof domain (protein family ID: PF02701) to query the Dof genes in the P. trichocarpa genome. All the obtained Dof genes were checked using the SMART database to ensure the presence of the Dof domain. Then, 41 Dof genes were identified and was close to results from various biology analyses which estimated 30, 36, 24, 31 and 54 Dof genes in rice, Arabidopsis, barley, wheat and maize, respectively [34][35][36].
Molecular masses of the Dofs in P. trichocarpa varied from 17735.9 to 55263.5 Da. The encoded proteins ranged from 159 to 1485 amino acids (aa), with an average of 684 aa. The location of the protein expression levels in the plant cell were predicted with WoLF PSORT. The majority Dof genes were predicated as nuclear proteins, except PtrDof23, which was located in the mitochondrial matrix (Table 1).

Phylogenetic analysis, exon/intron structure analysis and identification of conserved motifs
To examine the phylogenetic relationships, an unrooted phylogenetic tree was drawn ( Fig 1A). The 41 PtrDof homologs were separated into four groups (A to D). Subgroup A and D constituted the largest clade with 12 members each, subgroup B contained the fewest PtrDof family members (8 genes), and group C contained 9 members.
To check the structural diversity, we investigated the characterization of exon-intron structure in the genomic DNA sequences of individual PtrDof genes ( Fig 1B). The predicted numbers of exons among the PtrDof genes were relatively fewer, varying from one to three, with 19 members having one and 19 with two. Three (PtrDof14, PtrDof39 and PtrDof11) genes had three exons. Furthermore, some PtrDof genes belonging to the same subgroup had similar gene structures, such as introns numbers and extons lengths. For instance, subgroup A genes had one or two introns expect PtrDof33, and subgroup C genes had zero introns with exception of PtrDof17 and PtrDof 4. The majority of three-exon-genes belonged in subgroup A. These similar structural features may be related to their functions in the Populus genome ( Fig 1B). MEME programmer was used to assess PtrDof proteins to characterize motif compositions ( Fig 1C). A total of 15 conserved motifs were identified (S3 Table). These motifs are shown in the location of corresponding protein. The majority of related members in the phylogenetic tree had motifs. All genes uniformly contained common motif 1 at the C-terminal region, which was confirmed to be a conserved Dof domain. Moreover, differences in gene composition and motif organization among related PtrDofs members within the same subgroups indicated that these genes may have a divergent function.

Chromosomal location and gene duplication
In silico mapping of the gene loci indicated that the distribution of PtrDof genes span all 19 linkage groups (LGs) in an uneven manner. For example, PtrDof3 was located in the chromosomal scaffold that was not marked in the figure. As shown in Fig 2, the 41 PtrDof genes were dispersed on 18 of the 19 Populus chromosomes (none in chromosome XVIII). Chromosome IV and XI harbored the most (4 of 41) genes. On contrast, only one gene was found on each of chromosome IX, XIII, XVI, XVI I, and XIX. The rest of the chromosomes harbored two or three genes. Researches previously showed that the Populus genome has gone through three circles of genome-wide duplication at any rate including multiple segmental duplications, tandem duplications and transposition events in that order [37]. The segmental duplication associated with the salicoid duplication event that happened 65 million years ago promoted the expansion of numerous multigene families [27,[38][39][40]. We mapped the PtrDof genes to the duplicated blocks based on the previous research to check sure the possible relationship between the Dof genes and segmental duplications. The distribution of genes associated with the corresponding duplicate blocks is demonstrated in Fig 2. Approximately 49% (20 of 41) genes were firstly located in duplicated regions. Twelve duplicated genes (PtrDof1, 6, 8, 10, 11, 12, 13, 23, 25, 26, 28 and 31) were only contained in one of the blocks and lacked duplicates in the corresponding block. While, eight genes were located outside any duplicated blocks. The segmental duplication also occurred in subgroup C. These results suggest that the PtrDof genes likely originated from both segmental and tandem duplications.

Promoter cis-element analysis
Phytohormones such as salicylic acid (SA), jasmonic acid (JA), ethylene (ET), and ABA are involved in various processes throughout plants to accommodate abiotic stresses. We identified the putative cis-acting regulatory DNA elements by analyzing the promoter sequences of all PtrDof genes. In the PtrDof gene promoter region, many key cis-elements that were related to environmental stress signal responsiveness were identified, such as MBS (MYB binding site, involved in drought-inducibility), HSE (heat stress-responsive element), C-repeat/DRE (cold and dehydration-responsive element), TC-rich repeats (defense and stress-responsive element), LTR (low temperature-responsive element), and W-Box (WRKY binding site, involved in abiotic stress responsiveness). Other key elements included those in phytohormone signaling, such as ABRE (abscisic acid-responsive element), ERE (ethylene-responsive element), TCA-element (salicylic acid-responsive element), CGTCA-motif (MeJA-responsive element), TGACG-motif (MeJA-responsive element), and P-box (gibberellin-responsive element) (S4 Table). Most of the PtrDof genes containing cis-elements contained responsiveness to phytohormone signaling and environmental stress signal, whereas PtrDof18 and 41 contained only three cis-elements (S5 Table).

Gene Ontology (GO) annotation
The 41 PtrDof genes were classified into biological processes, molecular functions and cellular components by Gene Ontology (GO) using Blast2GO v3.0 (Fig 3, S6 Table). The seven terms   Table. of biological processes were defined. The function of all PtrDofs were predicated in the metabolic process, the cellular process and the biological regulation process, followed by the singleorganism (~7%) and the multicellular organismal processes and reproduction or developmental process (~5%). Molecular function predictions showed that all PtrDofs were in accordance with organic cyclic compound binding and heterocyclic compound binding. In addition, some PtrDofs were annotated as sequence-specific DNA binding involved in transcription factor activity (~17%). Furthermore, cellular component prediction indicated that four PtrDofs were localized in the membranes (~10%). Only two PtrDofs were organelle-localized and celllocalized (~5%).

Expression profiles under drought stress
A majority of land plants encounter environmental stress during their life span and drought is the major environmental stress. In order to characterize the possible roles of PtrDof genes in drought stress, we used the publicly available data to investigate the expression profiles of PtrDof genes responded to drought stress. A detailed description of gene expression was downloaded from the exPlot tool at the PopGenIE v3 database. The data for all the PtrDof genes are shown in S7 Table, except for PtrDof5, 11,18, and PtrDof3, for which the root has no data. 31 genes were upregulated in leaves and 14 genes were upregulated in roots under drought stress (Fig 4). Ten genes were all upregulated in leaves and roots.

Expression pattern of PtrDof genes under osmotic and ABA stress
Numerous PtrDof genes were obtained by heatmap analysis with differences in the expression profiles under drought stress. To verify these results, qRT-PCR was used to study the differential expression of selected genes under osmotic stress. For the leaves, 24 genes were induced, 4 genes were suppressed, and 3 genes were irregular (Fig 5). The results were broadly consistent with the heatmap data, which means the drought-induced genes were all upregulated and the drought-suppressed were all downregulated in leaves. For the roots, 15 genes were induced and 18 genes were suppressed (Figs 6 and 7). A total of 15 genes (PtrDof1, 8, 10, 17, 20, 21, 22, 24, 29, 32, 34, 36, 38, 40 and 41) were significantly downregulated at all time points, while only 8 genes (PtrDof14, 16, 19, 25, 27, 28, 37 and 39) were upregulated at all time points. The expression tendencies were roughly consistent with the exPlot analysis. In addition, PtrDof14, 16,25,27,28,37 and 39 were all upregulated in the leaves and roots under osmotic stress (Figs 5, 6 and 7).

MiRNA target site prediction and validation
Among the PtrDof genes, 12 were targeted by 15 miRNAs. PtrDof12, PtrDof13 and PtrDof40 were targeted by two miRNAs (S8 Table). PtrDof12 was targeted by ptc-miRf11023-akr and ptc-miRf12020-akr. PtrDof13 was targeted by ptc-miRf10053-akr and ptc-miRf10957-akr. PtrDof40 was targeted by ptc-miRf10540-akr and ptc-miRf11148-akr. The other PtrDofs were only targeted by one miRNA. The sequence analysis of PtrDof30 implied that the 1170-1190 bp region may be the target site of ptc-miR472b. Ptc-miR472b miRNA was previously    implicated with stress response [41]. PtrDof 30 was confirmed as a real target of miRNA, as all of the 5 0 ends of the mRNA fragments mapped to the nucleotide that paired to the tenth nucleotide of each miRNA with higher frequencies than depicted for each pairing oligomer (Fig 9).

Discussion
The Dof genes are ubiquitous in many plant species and are specific plant transcription regulators that are involved in various abiotic stress responses. The functional and evolutionary analysis of Dof genes have been preliminary performed in Arabidopsis, rice, soybean, Chinese cabbage, potato, pigeonpea, tomato and cucumber. In this report, we conducted a comprehensive analysis of the PtrDof family in P. trichocarpa to confirm their potential functions in response to osmotic stress.
We identified 41 putative full-length PtrDof genes in the P. trichocarpa genome. The number of PtrDof homologues was similar to that in Arabidopsis and rice [42]. The lengths of these sequences varied, which implied a highly complicacy among the PtrDof genes. Approximately 98% (40 of 41) of the PtrDof genes were identified to be localized in the nuclear, however only one gene was identified as the mitochondrial matrix (Fig 1A). The results were same as a previous analysis of the Dof gene family in cucumber where CsDof proteins were predicted as the nucleus except for CsDof30, which was extracellular [43].
Similar to previous discoveries in rice and Arabidopsis [42], the PtrDof genes had few introns (0-2) in each gene (Fig 1B). The motif analysis showed that motif 1 was uniformly observed in all Dof proteins (Fig 1C), similar to Arabidopsis, rice and tomato. This result indicated that the evolution of PtrDof transcription factors was conserved in plant development.
With high contribution to genetic novelty, the whole-genome, tandem and segmental duplications are important for genomic expansion, as the fundamental sources of genetic novelty [44][45]. We observed that 78% of PtrDof genes appeared to contain duplicated regions; however, eight genes were located away the duplicated blocks (Fig 2). The results suggested that partial genomic duplication included rearrangements, which resulted in the loss of a number of genes. The present study found only one pair (PtrDof4/34) with tandem duplication and ten pairs of segmental duplication events in PtrDof genes. This indicated that PtrDof gene segmental duplication and not tandem duplication is predominantly involved in the evolution of P. trichocarpa. The result was similar to that observed for CsDof duplications in cucumber [43].
The cis-elements play vital parts in the transcriptional regulation of gene expression, controlling phytohormone responses and complicated abiotic stress, to increase the resistance of plants under fluctuating environments. In our report, many cis-elements related to abiotic stress and phytohormone have been identified, including ABRE, MBS, HSE, ERE and TCAelements (S4 Table). Every gene in the family included three cis-elements mentioned at least (S5 Table). PtrDof18 and 41 have only three cis-elements, which indicated that these genes might not be related to abiotic stress. By comparison, PtrDof11, 16,19 and 36 have nine cis-elements, which indicated that these genes might be strongly associated with functions under different abiotic stresses. According to the expression profiles of the Populus Dof genes under various stresses, PtrDof16, PtrDof19 and PtrDof36 were upregulated in leaves at some time points after osmotic and ABA treatment. These results were roughly consistent with the promoter analysis.
The Dof family of transcription factors is a major large class of plant-specific factors, which have been identified to be involved in the regulatory networks of plants as very critical roles in response to abiotic stress. The exPlot data indicated that PtrDofs have important functions in drought stress. Under osmotic conditions, a total of 7 genes (PtrDof14, 16, 25, 27, 28, 37 and 39) in the Populus Dof gene family were all upregulated at all time points, including short and long term, in the leaves and roots (Figs 5, 6, 7 and 8). These results further suggest the involvement of these genes in osmotic stress as well as long-term response genes to osmotic stress. An Arabidopsis homolog (AT5G66940) of two of the upregulated genes (PtrDof27 and 28) was similarly upregulated under drought conditions [46]. Under ABA conditions, PtrDof27 and 28 and their homolog gene (AT5G66940) also have similar responsive expression in the leaves and roots, which were both induced [46]. Moreover, in response to osmotic treatment, osmotic upregulated PtrDof members grouped into four subgroups in the leaves. In Chinese cabbage, the 9 BraDof genes from different classes were all induced by drought stress in the leaves [4]. However, in Triticum, only two TaDof genes from two different clades were significantly upregulated by drought [36]. Probably, the mechanisms of Dof genes in response to abiotic stress may be various in different plant species [4]. We found that eight genes were upregulated under osmotic and ABA treatments in roots. The results indicated that PtrDofs may be involved in the development of plant roots. A previous study found that Populus Dofs were associated with the formation of adventitious roots [47]. Different PtrDof genes had different expression pattern under the same conditions. Our research illustrates that PtrDof transcription factors act as a vital role resistance to osmotic stress in plants.
The miRNAs act as a vital role during plant development and abiotic stresses-response. We found that 12 PtrDof genes were targeted by 15 miRNAs. Each miRNA acts on different genes in most cases. It was indicated that 16 miRNA families could target 13 CmDOFs previously and there was no repeat among these miRNAs [43]. The results indicated that each miRNA has its specific regulatory genes and that one miRNA may not simultaneously act on different Dofs. Four genes, including PtrDof12, PtrDof13 and PtrDof40, were targeted by two miRNAs, whereas the other nine genes only were targeted by one miRNA. Among the Dof gene family in the cucumber, there were three target sites for CmDOF1 and two target sites for CmDOF21, whereas the remaining 11 CmDOFs have only one target site [43]. In plants, single or multiple miRNAs are induced to regulate the expression of target genes under various stresses to improve the adaptability of plants [48]. The 12 PtrDof target genes belong to four groups of PtrDofs. Subgroup D contains the most target genes (four), whereas subgroup B only contains one target gene. Subgroup A and C, respectively, contain four and two target genes. In the cucumber, each subgroup also has a different number of target genes [43]. MiR472 is a miRNA induced by different stresses. The recent study showed that miR472 acts as a negative regulator, preventing an autoimmune response that would have detrimental consequences on plant fitness in Arabidopsis [49], and the targets of miR472 were identified as disease resistance genes in Arabidopsis lyrata [50]. The expression of miR472 plays a role in influencing the expression of their target genes, which are involved in disease resistance in Citrus sinensis [51]. The ptc-miR472b miRNA is regulated by cold stress in Populus [41], but no definite results are available. The expression of miR472 has been shown in the PMRD and the expression in rice was decreased under oxidase stress. The gene expression omnibus of NCBI has reported that miR472 had different expression in an experiment designed by Zhongs et al (http://www.ncbi. nlm.nih.gov/geo/query/acc.cgi?acc=GSE11535). However, to date, the interactions between miR472 and osmotic stress-related targets have not been reported. Our study provides a new target gene, PtrDof30, which is a osmotic-related gene in P. trichocarpa. We also confirmed that PtrDof 30 was a real target of miRNA in plants (Fig 9).

Conclusions
Dof-family transcription factors were comprehensively analyzed in the genome of Populus trichocarpa. We performed 41 PtrDof genes that were classified into four groups. Most genes within the same group had similar gene structures and conserved motifs. PtrDofs were dispersed on 18 of the 19 Populus chromosomes and likely originated from both tandem and segmental duplications. Promoter cis-element analysis indicated that most PtrDof genes contain cis-elements in response to stress and phytohormones. The HeatMap data for the PtrDof genes suggest that they are primarily expressed in leaves and roots. The expression of selected PtrDof genes was characterized in response to osmotic and ABA stresses and indicated that PtrDofs may be involved in resistance to abiotic stress in Populus trichocarpa. In addition, 12 PtrDof genes could be the targets of 15 miRNAs and PtrDof30 was confirmed as a real target of miRNA. The results help to characterize the stress responses of PtrDof genes and promote a better understanding of the construction and function of Dofs in Populus.
Supporting Information S1