Genome-Wide analysis of the AAAP gene family in moso bamboo (Phyllostachys edulis)

Background Members of the amino acid/auxin permease (AAAP) gene family play indispensable roles in various plant metabolism and biosynthesis processes. Comprehensive analysis of AAAP genes has been conducted in Arabidopsis, rice, maize and poplar, but has not been reported from moso bamboo. Phylogenetics, evolutionary patterns and further expression profiles analysis of the AAAP gene family in moso bamboo (Phyllostachys edulis) will increase our understanding of this important gene family. Results In this current study, we conducted phylogenetic, gene structure, promoter region, divergence time, expression patterns and qRT-PCR analysis of the 55 predicted AAAP genes in moso bamboo based on the availability of the moso bamboo genome sequence. We identified 55 putative AAAP (PeAAAP1-55) genes, which were divided into eight distinct subfamilies based on comparative phylogenetic analysis using 184 full-length protein sequences, including 55 sequences from moso bamboo, 58 sequences from rice and 71 sequences from maize. Analysis of evolutionary patterns and divergence showed that the PeAAAP genes have undergone a extensive duplication event approximately 12 million years ago (MYA) and that the split between AAAP family genes in moso bamboo and rice occurred approximately 27 MYA. The microarray analysis suggested that some genes play considerable roles in moso bamboo growth and development. We investigated the expression levels of the 16 AAP subfamily genes under abiotic stress (drought, salt and cold) by qRT-PCR to explore the potential contributions to stress response of individual PeAAAP genes in moso bamboo. Conclusions The results of this study suggest that PeAAAP genes play crucial roles in moso bamboo growth and development, especially in response to abiotic stress conditions. Our comprehensive, systematic study of the AAAPs gene family in moso bamboo will facilitate further analysis of the functions and evolution of AAAP genes in plants. Electronic supplementary material The online version of this article (doi:10.1186/s12870-017-0980-z) contains supplementary material, which is available to authorized users.

members of amino acid transporters were located within the tonoplast, which were devoted to transport amino acids between vacuole and cytoplasm, and regulated the storage of amino acids in vacuole [11][12][13][14].
To date, some functions of AAAP proteins have been studied in model plants such as Arabidopsis [20], poplar [21], maize [22] and rice [23]. AtAAP3 appears to be involved in amino acid uptake from the phloem and soil [24]. A recent study showed that AtAAP5 plays a role in amino acid uptake by the root [1]. AtAAP6 is expressed in roots, sink leaves, cauline leaves and xylem parenchyma, suggesting that it functions in amino acid uptake from the xylem [25]. In addition, AtAAP8 might play a crucial role in amino acid transport during fruit development [1,26]. In rice, 18 genes in the AAP subfamily have been identified [23], three of which (OsAAP1, OsAAP7 and OsAAP16) encode general AAAP proteins, whereas OsAAP3 does not [27]. OsAAP3 transports the basic amino acids lysine and arginine and has distinct substrate specificity compared with other rice or Arabidopsis AAPs [27]. OsAAP6 is contribute to enhance root absorption and affect the distribution of various amino acids in early stages of seed development [28].
Bamboo, one of the most important non-timber forest products worldwide, comprises over 70 genera and 1,200 species [29]. A majority of these species are distributed in the subtropical regions of China, especially regions south of the Yangtze River. Moso bamboo is an important species in China with the highest value in several areas among all bamboos, being used to produce timber, paper, artwork and food (young shoots) [30]. However, moso bamboo faces many types of environmental conditions during growth and development, such as high or low temperatures, salt concentrations and soil moisture levels, which limit its distribution and quality. A previous study showed that functional and regulatory proteins contribute to abiotic stress resistance in plant [31], and AAAP proteins are the fundamental functional proteins. Therefore, in the current study, we investigated AAAP proteins in moso bamboo to identify proteins that function in stress resistance. To date, bioinformatic analysis in model plants has greatly increased our understanding of AAAP genes. In addition, the draft genome sequence of moso bamboo was completed in 2013 [29], providing a great bioinformatics foundation to perform a comprehensive genome survey of the AAAP family in moso bamboo.

Identification of moso bamboo AAAP genes
The conserved AAAP domains (PF01490) of rice AAAP protein sequences were originally applied as seed sequences to search the NCGR database (www.ncgr.ac.cn/bamboo) [29]. Redundant sequences were removed manually based on the results of Cluster W 2.11 alignment [32], and each candidate sequence was confirmed using the Pfam (http:// pfam.xfam.org/) [33,34] and SMART (http://smart.emblheidelberg.de/) databases [35]. The number of amino acids, CDS lengths and physicochemical parameters of AAAP genes were obtained from Bamboo GDB (http://www.bamboogdb.org). Comparing coding sequence and the corresponding genomic DNA sequences of AAAP genes, we obtained their exon/intron structures from GSDS. The TMHMM Server version 2.0 (http://www.cbs.dtu.dk/services/TMHMM/) was used to predict the putative TM (transmembrane) regions of each PeAAAP protein with default settings.

Phylogenetic and conserved motif analyses
Multiple sequence alignment was performed using ClustalX 2.11 software [36], and a phylogenetic tree was constructed based on the alignment with the N-J method using MEGA 6.0 software and bootstrap analysis of 1,000 replicates. The combined phylogenetic tree of OsAAAP, ZmAAAP and PeAAAP proteins was generated using the same method. The motifs of PeAAAP proteins were identified using the MEME tool (http://meme-suite.org/tools/meme) (parameter setting: maximum number of motifs, 20; maximum width, 50.).

Calculation of Ka/Ks values
Pairwise alignment of AAAP genes encoding sequences of the orthologous and paralogous pairs was first performed using ClustalX 2.11 software and the results of alignment were subsequently further analyzed using the MEGA 6.0, and then the synonymous substitution rate (Ks) and nonsynonymous substitution rate (Ka) were computed using DnaSP 5 software [37,38]. The divergence time (T) was calculated using the formula T = Ks/2λ (λ = 6.5 × 10 −9 ) [29,39]. The following parameters were used to perform sliding window analysis of the Ka/Ks ratios of all homologous gene pairs: window size, 150 bp; step size, 9 bp.

Putative promoter region analysis
The 2,000-bp upstream sequences of the genetic sequences were identified as putative promoter regions, which contains various cis-regulatory elements identified using the PLACE website (http://www.dna.affrc.go.jp/PLACE/) [40].

Plant material and growth conditions
Eight-week-old seedlings were grown in artificial growth chamber with a constant photoperiod (14 h light/8 h darkness) and temperatures average around 22°C. Moso bamboo seeds for breeding seedlings were collected in the Tianmu Mountain National Nature Reserve in Zhejiang Province, China. In addition, the permission of seeds collection for the experiments was obtained from Prof. Dingqing Tang of School of Forestry and Bio-technology, Zhejiang A & F University. And the identification of these seeds was also performed by Prof. Dingqing Tang. The seedlings were treated with 20% PEG-6000, 200 mM NaCl and 4°C to induce drought stress, salt stress and cold stress, respectively. In order to obtain reliable experimental data and reduce experimental error, for each sample, we executed three repeated trials for the same stimulation and carried out three biological replicates for expression analysis. For each induction treatment, we collected samples at six time points (0, 1, 3, 6, 12 and 24 h) and immediately stored at -80°C freezer for RNA extraction. In addition, untreated plant materials (0 h) were used as the control group.

Expression profile analysis
To study gene expression levels of PeAAAP genes in different tissues or development stages. The expression profile for each gene was obtained from Short Read Archive (SRA) database of NCBI. And then the raw RNA-seq reads of BioProject ERP001341 were trimmed to remove low quality base-calls (Q < 20) and adaptor sequences with pipeline Fastq clean [41]. The paired clean reads were mapped to the Phyllostachys heterocycla reference genome using pipeline tophat2 with defaults parameters, and different expressed genes were detected by Cufflinks [42]. The heatmap of PeAAAP genes in seven different tissues and/or developmental stages (leaf, early panicle, advanced panicle, root, rhizome, 20-cm shoot and 50-cm shoot) was exhibited using the Heatmapper Plus tool [43].

qRT-PCR analysis
To research the expression levels of PeAAAP genes, qRT-PCR analysis based on SYBR-green fluorescence was performed for each members of the AAP subfamily. Total RNA was extracted from the plant samples using RNA prep Pure Plant Kit (Tiangen) according to the manufacturer's instructions, which was reverse transcribed into cDNA subsequently using a PrimeScript™ RT Reagent Kit (TaKaRa). Primer Express 3.0 was used to design the gene-specific primers of each PeAAP genes, and the tonoplast intrinsic protein 41 (TIP41) was used as an internal control [44]. The following program was used for qRT-PCR: 95°C for 30 s; 40 cycles of 95°C for 10 s, 55°C for 15 s, 72°C for 10 s.

Identification of AAAP genes in moso bamboo
The AAAP candidate sequences from the moso bamboo genome were verified using the Pfam (http://pfam.xfam.org/) [33,34] and SMART (http://smart.embl-heidelberg.de/) databases [35]. Fifty-five potential AAAP sequences were ultimately identified as AAAP genes, which were designated PeAAAP01 to PeAAAP55. Detailed information about the 55 AAAP genes was obtained using the moso bamboo GDB server (http://www.bamboogdb.org), including the predicted lengths of CDSs, sizes of encoded proteins and physicochemical parameters; this information is shown in Table 1. The lengths of the CDSs range from 252 bp to 2,166 bp, with an average size of 1,281 bp. The identified AAAP genes in moso bamboo with an average size of 426 aa, which peak on 721 aa and have a minimum value at 83 aa. The predicted molecular weights of the 55 PeAAAP gene products range from 9.28 kDa (PeAAAP3) to 77.56 kDa (PeAAAP10), with a mean value of 46.48 kDa. The predicted PIs for the 55 PeAAAP gene products are below 11.0, with most values approximately 8.0 or 9.0. However, the PI of one gene product (PeAAAP29) is below 5.0, whereas one is greater than 10.0 (PeAAAP37).
The GSDS online tool was used to determine the exons/ introns structure of each predicted PeAAAP gene (Fig. 1). These results show that five PeAAAP genes (PeAAAP4, PeAAAP16, PeAAAP43, PeAAAP48 and PeAAAP51) exclude introns, while the remaining genes contain 1-11 introns, and PeAAAP2 is no upstream and downstream (Fig. 1). The TMHMM Server v2.0 was used to predict the putative TM regions of PeAAAP genes, which numbers in most genes ranged from 8 to 13 (Additional file 1: Figure S1). Interestingly, we found that genes of the same subfamily have similar numbers of TMs, especially all members of AUXs contain 10 TMs. In AAP subfamily, the numbers of TMs was less than or equal to 10. However, a previous report, showing that AtAAP1 had 11 TMs [45].

Phylogenetic and conserved domain analysis of AAAP proteins in moso bamboo
The phylogengtics tree was constructed based on the alignment of full-length amino acid sequences to evaluate the evolutionary relationship among these members of AAAP family. The genes were divided into eight distinct subfamilies, and the AAP subfamily is the largest one of all. To further examine the diversification of the PeAAAP genes, 20 distinct motifs (Fig. 2) were found using MEME web server (http://meme-suite.org/tools/meme). Detailed information about the 20 putative motifs included names, widths and best possible matches were listed in Additional file 2: Table S1. The functions of each motif were identified by searching Pfam and SMART database, showing that nine motifs (1 -6, 9, 10 and 15) encode Aa-trans domains, while the remaining 11 motifs do not encode any domain. Interestingly, some motifs were specific to only one or two subfamilies. For example, motifs 8, 10 and 13 are exclusively found in the AUX subfamily, and motif 1 is only present in the AAP subfamily. Motifs 2 and 4 are found in all of the subfamilies except the ATLa and PorT subfamily, respectively. All members of the AUX subfamily have similar numbers of motifs (Fig. 2), indicating that the structures of members are highly conserved in the same subfamily. Evolutionary patterns and divergence of the AAAP gene family in moso bamboo, rice and maize We identified 14 paralogs (Pe-Pe) in the moso bamboo genome and 21 orthologs (Pe-Os) between moso bamboo and rice, and 19 orthologs (Pe-Zm) between moso bamboo and maize, using phylogeny-based and bidirectional best-hit methods. The formula T = Ks/2λ was used to evaluate the divergence times between moso bamboo, maize and rice, and relative Ks values were used as a proxy for time. All of the paralogous and orthologous pairs are listed in Table 2. The distribution of Ks values of paralogous pairs (Pe-Pe) peaked at approximately 0.15, as shown in Fig. 3. This result suggests that AAAP genes in moso bamboo have undergone a large-scale duplication event approximately 12 million years ago (MYA). The Ks values distribution of orthologous genes (Pe-Os and Pe-Zm) is shown in Fig. 4a and Fig. 4b, respectively. Interestingly, both Pe-Os and Pe-Zm had the same peak at approximately 0.35, showing that these genes have diverged approximately 27 MYA. A previous study showed that bamboo underwent whole-genome duplication 7-12 MYA, while rice and moso bamboo, as well as maize and moso bamboo, diverged 48.6 and 64.6 MYA, respectively [29]. It indicated that the AAAP family has undergone gene evolution after its separation from rice and maize. In general, Ka/Ks ratio less than 1, equal to 1 and greater than 1 means negative or stabilizing selection, neutral selection and positive selection, respectively [46,47]. To investigate the trend of Ka/Ks ratio in the coding sequences of each gene pairs, we performed sliding-window analysis of Pe-Pe (Additional file 3: Figure S2), Pe-Os (Additional file 4: Figure S3) and Pe-Zm (Additional file 5: Figure S4) gene segments, indicating that the AAAP domains have undergone strong purifying selection (Ka/Ks < < 1) during the process of evolution.

Putative promoter region analysis
Cis-elements play critical roles in plant growth and development, including determining the tissue-specific or stress-responsive expression patterns of genes, and multi-stimulus-responsive genes are closely correlated with cis-regulatory elements in their promoter regions [48,49]. Cis-elements have decisive effects on binding to target genes. In this study, we identified three type cis-elements, including cold-responsive, drought-responsive and salt-responsive elements in the promoter regions to help elucidate the potential functions of AAAP genes in moso bamboo [50,51]. Numerous cis-elements were widespread in the promoter regions, such as S000176 and S000415 for drought stress, S000453 for salt stress and S000407 for cold stress (Additional file 6: Table S2). Moreover, contrasting with the cis-regulatory elements of salt stress, there was the higher amount of cis-elements for drought and cold stress. These results suggest that transcription factors that regulate AAP genes may respond to abiotic stress and have the potential for improving abiotic stress responses, especially drought and cold. These findings may be helpful for further investigating stress tolerance mechanisms in moso bamboo.
Comparative analysis of AAAP genes in moso bamboo, rice and maize To date, most studies have focused on analyzing the AAAP family in rice and maize. To further analyze the evolutionary relationships between AAAP genes in moso bamboo, rice and maize, we constructed an N-J phylogenetic tree (Fig. 5) of AAAPs using ClustalX 2.11 based on 184 full-length AAAP protein sequences, including 55 sequences from moso bamboo, 58 from rice and 71 from maize. The detailed characteristics of the AAAP genes from rice and maize are listed in Additional file 7: Table S3. The phylogenetic tree clearly shows that the 184 AAAP proteins could be divided into eight distinct groups. A count of the number of AAAP proteins in every moso bamboo (Fig. 6b), rice (Fig. 6c) and maize (Fig. 6d) subfamily was performed. The result is consistent with the previous reports that AAP subfamily is the largest one in AAAP family among these three species.
We also found that moso bamboo contains fewer AAAP genes than rice and maize (Fig. 6a), suggesting that the PeAAAP genes have undergone a gene loss event after divergence from the last common ancestor of moso bamboo, rice and maize.

Differential expression profiling of moso bamboo AAAP genes
In general, the overall analysis of gene expression profiles in different tissues will contribute to study the dynamic gene expression of AAAP genes in moso bamboo. The highthroughput RNA sequencing (RNA-Seq), as one of essential next generation sequencing technology, will allow to reveal a snapshot of RNA presence and quantity from a genome at a given moment in time [52,53]. In addition, the draft genome sequence of moso bamboo has been released [29]. By now, many studies of expression profiles in several gene families were reported and mainly focused on different tissues [54][55][56]. While, the genome-wide expression profile of PeAAAP genes still remains unclear. We performed a microarray analysis to determine the expression level of each AAAP genes on different tissues in moso bamboo based on the above advantages. Finally, the heatmap was produced (Fig. 7) based on the microarray data of 55 moso bamboo AAAP (Additional file 8: Table S4) genes downloaded from the NCBI. From the microarray results, it was apparent that a few AAAP genes exhibited tissue-specific expression patterns. For instance, two genes (PeAAAP1 and PeAAAP5) and PeAAAP34 were highly expressed in advanced panicle and leaves, respectively. While the rest members of PeAAAP genes showed express at least two tissues. Above all, there are twenty-one genes (PeAAAP7, PeAAAP10, PeAAAP13, PeAAAP14, PeAAAP15, PeAAAP16, PeAAAP17, PeAAAP27, PeAAAP28, PeAAAP29, PeAAAP36, PeAAAP41, PeAAAP43, PeAAAP45, PeAAAP50, PeAAAP51 and PeAAAP53) that were widely expressed in all these seven tissues or developmental stages, implying their essential roles in the process of moso bamboo growth and development.
Six gene pairs (PeAAAP12/PeAAAP41, PeAAAP14/PeA AAP36, PeAAAP21/PeAAAP38, PeAAAP26/PeAAAP49, PeA AAP43/PeAAAP55 and PeAAAP44/PeAAAP52) of these above identified exhibit distinct expression patterns in different tissues or developmental stages, suggesting that duplicated genes may have different evolutionary fates. For instance, PeAAAP43 is expressed at a high level in rhizome and shoots, however, its counterpart PeAAAP55 shows slight relative expression level. By contrast, remaining gene pairs have the same or similar patterns of expression accumulation.

qRT-PCR analysis of moso bamboo AAAP genes
The phylogenetic analysis indicated that the AAP subfamily contains 16 PeAAAP genes and that these genes are closely related to stress-responsive genes in rice. This observation prompted us to investigate possible stress-responsive genes among the 16 PeAAAP genes by qRT-PCR. We investigated the expression levels of 16 selected AAP subfamily members (PeAAAP1, PeAAAP5, PeAAAP9, PeAAAP11, PeAAAP14, PeAAAP17, PeAAAP18, PeAAAP21, PeAAAP25, PeAAAP26, PeAAAP33, PeAAAP34, PeAAAP36, PeAAAP38, PeAAAP40 and PeAAAP49) in the leaves of young seedlings in response to PEG, salt and cold treatment using qRT-PCR. The specific primers used in qRT-PCR analysis of these genes are shown in Additional file 9: Table S5.
Two genes in the AAP subfamily (PeAAAP11 and PeAAAP49) were downregulated by cold stress, whereas 13 other genes of this subfamily (PeAAAP5, PeAAAP9, PeAAAP14, PeAAAP17, PeAAAP18, PeAAAP21, PeAAAP25, (See figure on previous page.) Fig. 1 Phylogenetic relationship and gene structures of AAAP proteins in moso bamboo. Left: Phylogenetic tree of PeAAAPs constructed by neighbor-joining method based on the results of sequence alignment. Bootstrap values from 1000 replicates were indicated at each node. The proteins on the tree can be divided into eight distinct subfamilies and the branches of different subfamilies were marked by different colors. There was a special noted that subfamilies ATL were further divided into two groups (ATLa and ATLb). Right: Exons, introns and untranslated regions (UTRs) were indicated by yellow rectangles, gray lines and blue rectangles, respectively Fig. 2 Schematic representation of the 20 conserved motifs in PeAAAP proteins. Conserved motifs of the PeAAAP proteins were identified using the online MEME program based on 55 full length amino acid sequences. Each specific motif was marked by different colored box, and their names in the center of the boxes. The length of each box in the figure didn't represent the actual motif size PeAAAP26, PeAAAP33, PeAAAP34, PeAAAP36, PeAAAP38 and PeAAAP40) were clearly upregulated by this treatment. Six genes (PeAAAP9, PeAAAP18, PeAAAP33, PeAAAP34, PeAAAP38 and PeAAAP40) were the most highly expressed during early (1 h) treatment, and their expression gradually decreased at all later time points (Fig. 9). Moreover, these genes were dramatically upregulated at the 1 h time point, especially PeAAAP9, PeAAAP34 and PeAAAP38 (more than 150-fold that of control levels; Fig. 9). The expression of PeAAAP1 changed only slightly over the 24-h time course. Three genes (PeAAAP5, PeAAAP21 and PeAAAP25) exhibited similar trends in expression, with a gradual increase in expression during the early time points, a peak at 6 h and a significant, gradual decrease at all later time points. The expression of PeAAAP14, PeAAAP26 and PeAAAP36 peaked at 3 h, while the expression level of PeAAAP17 was highest at 12 h (Fig. 9).

Discussion
As an indispensable gene family, the eukaryotic-specific AAAP family plays a pivotal role in the process of plant growth and development, which is one of the largest families of AATs identified to date [21,22]. According to previous studies, several AAAP genes have been characterized both physiologically and genetically, such as Arabidopsis AAP and AUX subfamily [18,25,57,58], and rice AAP subfamily [27,28]. By contrast, AAAP family members have not previously been characterized in moso bamboo. Therefore, in the current study, we identified and characterized 55 predicted AAAP genes in moso bamboo using genome wide analysis, and compared these with 58 OsAAAPs and 71 ZmAAAPs, showing that the number of AAAP genes in moso bamboo (55) is the fewest among these three species [22,23]. The 184 AAAP proteins were found to be divided into eight distinct subfamilies, with every subfamily containing different members from these three species, meaning that AAAP genes had diversified before moso bamboo, rice and maize split. A count of these subfamilies in different species shows that AAP subfamily is the largest one. And AAAP genes with the similar structure showed a tendency to be grouped into the same subfamily, we sought some more valuable evidence to support the reliability of the subfamily classification, including gene structure, motif compositions and TM regions. Furthermore, all members in AUX subfamily have the same numbers and types of TM regions, exon/intron structures and motif compositions, suggesting that these members share a closer evolutionary relationship in the process of AAAP evolution. These results are in accordance with the results of a previous study of the AAAP family in poplar [21]. Furthermore, these results may show Fig. 6 Comparison of AAAP proteins from moso bamboo, rice and maize. a: Comparison of AAAP proteins from moso bamboo, rice and maize; different color represents different species, and numbers of AAAP proteins in each species were marked. b, c and d showed that distribution and comparison of each subfamily for moso bamboo, rice and maize, respectively; different subfamilies were represented with different colors, and numbers of members in each subfamily were marked that the functions diversified among different members of AAAP genes in mso bamboo. In addition, a previous study showed that AAP1 had 11 TMs in Arabidopsis, which was different from AAP subfamily in moso bamboo, signifying the divergence between different species.
Recent gene duplication events, which help organisms adapt to different environments during growth and development [59,60] as well as are an important evolutionary mechanism for the rapid expansion and evolution of gene families [46]. To better explain the patterns of macroevolution in moso bamboo, we calculated the value of Ks and Ka in moso bamboo, maize and rice. Specifically, we estimated the Ks and Ka models of paralogous genes (Pe-Pe) and orthologous genes (Pe-Os and Pe-Zm) and calculated the Ks value for each gene pair. We estimated that a largescale duplication event was occurred approximately 12 MYA in moso bamboo and that the divergence times for orthologous genes (Pe-Os and Pe-Zm) was approximately 27 MYA. Peng et al. estimated that the divergence time between moso bamboo and rice was 48.6 MYA, and maize was 64.6 MYA [29]. Ratio of nonsynonymous to synonymous substitutions (Ka/Ks) can be used to measure the (See figure on previous page.) Fig. 7 Expression profiles of moso bamboo AAAP genes across different tissues and development stages. Heatmap showing hierarchical clustering of 55 PeAAAP genes across different tissues analyzed. Color scale erected vertically at the right side of the picture represents log10 expression values, green represents low level and red indicates high level of transcript abundances. L, leaf; P1, early panicle; P2, advanced panicle; R, root; Rh, rhizome; S1, 20-cm shoot; S2, 50-cm shoot. Every member of AAP subfamily was marked by red stars Fig. 8 Expression patterns of 16 selected AAP subfamily genes in moso bamboo under PEG treatment, as revealed by qRT-PCR. The Y-axis and X-axis indicates relative expression levels and the time courses of stress treatments, respectively. Error bars, 6 ± SE history of selection acting on coding sequences [61]. In general, Ka/Ks ratio less than 1, equal to 1 and greater than 1 means negative or stabilizing selection, neutral selection and positive selection, respectively [46,47]. Interestingly, in this study, the Ka/Ks ratios were less than 1, evidencing that the homologous gene pairs of AAAP family in moso bamboo have undergone a markedly purifying selection in the course of evolution.
The overall analysis of gene expression profiles in different tissues will contribute to study the dynamic gene expression of AAAP genes in moso bamboo. Therefore, we displayed the gene expression profiles of 55 identified PeAAAP genes using published transcriptome data in NCBI database. Among them, twenty-one genes (PeAAAP7, PeAAAP10, PeAAAP13, PeAAAP14, PeAAAP15, PeAAAP16, PeAAAP17, PeAAAP27, PeAAAP28, PeAAAP29, PeAAAP36, PeAAAP41, PeAAAP43, PeAAAP45, PeAAAP50, PeAAAP51 and PeA AAP53) exhibited relatively high expression level in all seven different tissues or developmental stages, suggesting their importance in the processes of moso bamboo growth and development. While a few numbers of AAAP genes show tissue-specific in this analysis.
In plant, many stress-related genes generated a series of stress responses to meet the adverse environmental condition during growth and development. AAAPs are highly regulated by environmental signals and play positive roles in abiotic stress responses in many plants [62,63]. AAP genes were previously studied in several species, leading to the identification of 18 and 8 genes in the AAP subfamily in rice and Arabidopsis, respectively. The expression pattern of a gene can provide significant clues about its function, as demonstrated by Zhao et al. [64]. Thus, we performed qRT-PCR to investigate the expression patterns of PeAAAP genes under PEG, cold and NaCl treatment. The PeAAP subfamily members showed significantly differential expression patterns under the three abiotic stresses examined. Most PeAAP genes were upregulated by all three abiotic stress treatments, suggesting that PeAAP genes may play crucial roles in abiotic stress responses in moso bamboo. For instance, PeAAAP9 was highly expressed (over 100-fold that of control levels) under PEG (drought), cold and salt stress treatment. However, this gene shows a relatively lower expression frequency in leaf, meaning that PeAAAP9 responds to environmental stress. And conversely, some genes are unresponsive, especially PeAAAP1 has low expression level in both microarray and qRT-PCR analysis. Furthermore, there were six paralogous pairs in AAP subfamily. Of these gene pairs, five of them (PeAAAP14/PeAAAP36, PeAAAP18/PeAAAP33, PeAAAP21/PeAAAP38, PeAAAP 25/PeAAAP40 and PeAAAP26/PeAAAP49) under cold treatment and two gene pairs (PeAAAP14/PeAAAP36 and PeAAAP18/PeAAAP33) under salt treatment (NaCl) had similar expression levels and tendency in the same paralogous pair. These results might suggest that homologous genes had similar putative functions in processes of organism growth and development.

Conclusions
In this study, we investigated phylogenetic, gene structure, promoter region, divergence time, microarray analysis and qRT-PCR analysis of the 55 predicted AAAP genes in moso bamboo. The qRT-PCR was used to explore the expression patterns of 16 selected AAAP genes under three abiotic stresses: drought (20% PEG-6000), salt (200 mM NaCl) and cold (cultured at 4°C). These results of this study increase our understanding of AAAP family members, including their possible contributions to abiotic stress responses and other putative functions in process of moso bamboo growth and development.