Figures
Abstract
Background
Phosphorus (P) is an essential macronutrient for plant growth and development. Several genes involved in phosphorus deficiency stress have been identified in various plant species. However, a whole genome understanding of the molecular mechanisms involved in plant adaptations to low P remains elusive, and there is particularly little information on the genetic basis of these acclimations in coniferous trees. Masson pine (Pinus massoniana) is grown mainly in the tropical and subtropical regions in China, many of which are severely lacking in inorganic phosphate (Pi). In previous work, we described an elite P. massoniana genotype demonstrating a high tolerance to Pi-deficiency.
Methodology/Principal Findings
To further investigate the mechanism of tolerance to low P, RNA-seq was performed to give an idea of extent of expression from the two mixed libraries, and microarray whose probes were designed based on the unigenes obtained from RNA-seq was used to elucidate the global gene expression profiles for the long-term phosphorus starvation. A total of 70,896 unigenes with lengths ranging from 201 to 20,490 bp were assembled from 112,108,862 high quality reads derived from RNA-Seq libraries. We identified 1,396 and 943 transcripts that were differentially regulated (P<0.05) under P1 (0.01 mM P) and P2 (0.06 mM P) Pi-deficiency conditions, respectively. Numerous transcripts were consistently differentially regulated under Pi deficiency stress, many of which were also up- or down-regulated in other species under the corresponding conditions, and are therefore ideal candidates for monitoring the P status of plants. The results also demonstrated the impact of different Pi starvation levels on global gene expression in Masson pine.
Citation: Fan F, Cui B, Zhang T, Qiao G, Ding G, Wen X (2014) The Temporal Transcriptomic Response of Pinus massoniana Seedlings to Phosphorus Deficiency. PLoS ONE 9(8): e105068. https://doi.org/10.1371/journal.pone.0105068
Editor: Enrique Hernandez-Lemus, National Institute of Genomic Medicine, Mexico
Received: February 19, 2014; Accepted: July 20, 2014; Published: August 28, 2014
Copyright: © 2014 Fan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The project is supported by grants from Special Core Program of Guizhou Province, P.R. China (20126011-1), the 863 Program of China (2011AA10020301), the Natural Science Foundation of Hubei Province P.R. China (2013CFC012) and the National Natural Science Foundation of China (31260183). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Phosphorus (P) is an essential mineral macronutrient that is crucial for plant development and metabolism. Plants acquire P in the form of inorganic phosphate (Pi) through their roots from water in the soil. Although Pi is generally abundant in soils, the availability for plant uptake is often limited due to high fixation and slow diffusion [1], [2], making P one of the major factors that constrain plant growth and productivity worldwide [2], [3]. Economically important crops and forestry plantations are therefore frequently supplemented with Pi fertilizer to maximize yields and quality. Unfortunately, the application of Pi fertilizer can cause severe environmental damage, such as eutrophication of water systems [4]. Pi fertilizer use also significantly increases the cost of crop production, particularly for forest cultivation, and worldwide Pi resources will be exhausted in the near future if the current rate of extraction continues [3], [4]. Thus, a fuller understanding of the strategies used by plants to uptake and utilize Pi is of high importance and may be applied to the breeding or engineering of plants with greater capacity to acquire, transport, store, and recycle soil Pi [5]–[7]. Plants have evolved a variety of adaptive strategies to cope with Pi-deficiency that involve in morphological, physiological, biochemical, and molecular responses. For example, plants may modify root architecture and morphology [8], [9], increase the exudation of H+, organic acids and phosphatases from roots [10]–[12], optimize internal Pi use [13], and elevate the expression of high-affinity phosphate transporters [14]–[16].
It is believed that these adaptive strategies are mediated by the coordinated function of hundreds of plant genes. Over the past two decades, extensive studies on the response to Pi starvation in the model plant Arabidopsis thaliana have contributed significantly to our understanding of P signaling and response pathways [17]–[19]. Meanwhile, transcriptional profiling of A. thaliana [20] and other important crop plants such as Oryza saliva (rice) [21], Zea mays (maize) [22], Brassica rapa (Chinese cabbage) [23], Cucumis melo (melon) [24], Phaseolus vulgaris (common bean) [25] and Lupinus albus (white lupin) [16] have provided knowledge at the whole genome scale and revealed the complexity of the network of genes necessary for plants to adapt to low soil Pi availability. A series of genes confer adaptation to Pi-deficiency, mainly through the regulation of P acquisition, internal remobilization, changes in metabolism, and signal transduction [26]. Expression of genes encoding purple acid phosphatases (PAPs) and Pi transporters were found to be generally up-regulated under Pi starvation, and these proteins are important for mobilization of soil organic P for root absorption and transportation [27], [28]. Pi-deficiency also up-regulated several genes involved in lipid metabolism, including nonspecific phospholipase and UDP sulfoquinovose synthases [29], and affected carbon metabolism [13]. Numerous regulatory genes involved in response to Pi-deficiency have also been identified, which included transcription factors (TFs), sugar-phosphate exchanger subfamily proteins (SPX, SYG1, Pho81, XPR1), plant hormones and protein modifiers, as well as noncoding RNAs [30]–[32].
Masson pine (Pinus massoniana), a gymnosperm belonging to the conifer genus, is native to southern China and is one of the most economically important forest trees that is widely used for timber, pulp and resin production. This tree species is grown mainly in tropical and subtropical regions, many of which are severely stricken with Pi-deficient soils, which constrains the productivity of this and other forest crops [33]. Several researches had been carried out to improve the P efficiency [34]–[37], however, the genes involved in the Pi starvation response in Masson pine have not yet been unraveled to date.
Gene expression microarray analysis may generate a wealth of important information on transcriptional changes in response to different stimuli, and have contributed enormously to many areas of the life sciences [38]. Nevertheless, microarray methods suffer some potential pitfalls, and RNA-seq approaches can provide superior reproducibility and accuracy, and can reveal transcriptional changes over a wider dynamic range. Thus, RNA-seq techniques are increasingly used to accurately quantify gene expression levels [39]–[42]. Kogenaru et al. (2012) showed that RNA-seq and microarray approaches can mutually complement demerits, and their combination may provide a more comprehensive understanding of the transcriptome [43].
An elite Masson pine germplasm, which showed superior tolerance to Pi-deficiency, has been obtained from our previous work. In order to investigate the complex molecular mechanisms involved, transcriptional profiling was performed by a combination of RNA-seq and microarray methodologies in this germplasm under Pi-sufficient and Pi-deficient conditions. The effects of the magnitude and duration of Pi-deficient conditions on global gene expression were investigated. The results provided insight into the molecular mechanisms involved in acclimation to long-term Pi starvation at different levels of Pi-deficiency in coniferous trees, and this knowledge may be applied to other plants. The differentially expressed genes identified in this study will facilitate the further genomic and genetic engineering studies on Masson pine and other important crops.
Materials and Methods
Plant materials and growth conditions
Masson pine seeds were collected from a first-generation orchard located in the Meishan Forest Farm of Sichuan Province (P. R. China) (E: 103.21; N: 29.25). These seeds were shown to have a high ability to survive in Pi-deficient conditions in our previous work. Field studies did not involve endangered or protected species, and no specific permissions were required for this location and activities. Seeds were surface sterilized, rinsed in sterile water, and allowed to imbibe overnight in a tray containing warm sterile water at 30°C. Sterilized seeds were planted for six seedlings per 12 centimetres diameter pot containing quartz sand. Pots were placed in growth chambers at 25°C/16°C with a 14-h photoperiod (2000Lx), and watered once per day since 10 days after emergence (DAE). Three treatments were carried out with different P levels: i) control nutrient solution containing 5.0 mM KNO3, 4.5 mM Ca(NO3)2·4H2O, 2.0 mM MgSO4·7H2O, 0.5 mM KH2PO4, 46 µM H3BO3, 10 µM MnCl2·4H2O, 0.8 µM ZnSO4·7H2O, 0.56 µM CuSO4·5H2O, 0.4 µM H2MoO4·4H2O, and 25 µM Fe-NaEDTA; ii) nutrient solution deficient in Pi (P1) containing 0.01 mM KH2PO4; iii) nutrient solution deficient in Pi (P2) containing 0.06 mM KH2PO4. KCl was added to Pi-deficient solutions to ensure identical potassium concentrations. Ten-day-old seedlings were divided into three groups, which were grown under one of the three P levels. The nutrient solution was added at a rate of 100 mL per pot every two days. Seedlings were harvested 22, 34, 58 and 70 DAE, immediately frozen in liquid nitrogen and stored at −80°C for RNA extraction. All treatments were replicated in three pots.
RNA isolation, cDNA library construction and sequencing
Total RNA was extracted from the samples using an Invitrogen Plant RNA Isolation Kit and purified using an RNeasy MiniElute Cleanup Kit (Qiagen), both according to the manufacturers’ protocols. RNA was quantified using a Qubit RNA Assay Kit (Invitrogen), and RNA integrity was checked with the RNA6000 Nano Assay using an Agilent 2100 Bioanalyzer (Agilent Technologies). RNA isolated from Pi-deficient samples (P1 and P2) and Pi-sufficient (control) samples were mixed separately for library preparation and sequencing. cDNA library preparation and deep sequencing were conducted in the Biochip National Engineering Research Center of Beijing, CapitalBio Corporation. For Illumina library construction and sequencing, RNA samples were prepared using a TruSeq RNA Sample Preparation Kit (Illumina) according to the manufacturer’s recommendations. The library was qualified using an Agilent 2100 Bioanalyzer and quantified by Qubit and qPCR assays. Cluster formation and sequencing were performed on a HiSeq2000 platform (Illumina) following the manufacturer’s standard cBot and sequencing protocols. Primary data analysis and base calling were performed using the Illumina instrument software.
Transcriptome assembly, functional annotation and classification
To generate a non-redundant set of transcripts, we performed an assembly with publicly available programs using 112,108,862 RNA-seq reads obtained from two libraries. High quality statistics were selected on the basis of Q20. The reads were assembled using Trinity software (Grabherr et al., 2011) to construct unique consensus sequences. Unigenes were compared with the NCBI non-redundant nucleotide database (Nt, January, 2013; http://www.ncbi.nlm.nih.gov/) and non-redundant protein database (Nr, January, 2013; http://www.ncbi.nlm.nih.gov/) using BLASTN and BLASTX, respectively, with E-value cutoffs≤1e-5. Putative functions were assigned to unique sequences by sequence similarity comparison against the Clusters of Orthologous Groups of proteins (COG) database [44], [45] with BLAST at E values≤1e-10. InterPro domains were annotated using InterProScan Release 27.0 and functional assignments were mapped onto the Gene Ontology (GO) entries [46]–[48]. Assembly sequence data was submitted to the NCBI Transcriptome Shotgun Assembly (TSA) database (accession number GAQR00000000).
Microarray preparation and data analysis
Unique 60-mer oligonucleotide probes were designed for each of the >56,000 high quality unigenes identified in this work. 8×60-K DNA microarray chips were implemented with the Agilent platform. A total of 12 sample RNAs consisting of one of each of the three different P levels from each experimental duration were further purified using an RNA Cleanup Kit (Macherey-Nagel, Germany). Total RNA was used with a cDNA Amplification Tag Kit (Jingxin, CapitalBio Corporation, China) and a NucleoSpinExtract II Kit (MN) to generate the targets. Each mRNA sample was reverse-transcribed in the presence of Cy3-dUTP or Cy5-dUTP (GE Healthcare) and 5 µg of labeled cDNA was hybridized with the microarray. After washing, microarray plates were dried briefly and scanned with an Agilent G2565CA Microarray Scanner. The ratios of the two fluorescent signal intensities of each DNA element were then measured to determine changes in gene expression. Microarray data were processed at the Biochip National Engineering Research Center of Beijing, CapitalBio Corporation. Analysis was performed using Feature Extraction and GeneSpring GX software with RMA normalization. Genes were considered differentially expressed if up-regulated by >2-fold or down-regulated by <0.5-fold (P<0.05). Microarray data have been deposited in the Gene Expression Omnibus database (accession number GSE52835).
Quantitative real-time PCR verification
Total RNA was isolated from seedlings grown under different Pi treatments 22, 34, 58 and 70 DAE as described above. First-strand cDNA was synthesized using an RNA LA PCR Kit (TaKaRa, Dalian, China) with the supplied oligo dT-adaptor primer following the manufacturer’s instructions. Gene-specific primers (Table 1) were designed for nine randomly selected unigenes involved in Pi-deficiency using Primer Premier 5.0. Samples and standards were performed in triplicate on each plate using the SYBR Select Master Mix Kit (Applied Biosystems) on a 7500 fast real-time PCR system (Applied Biosystems) following the manufacturer’s instructions. Quantitative real-time PCR (qRT-PCR) was performed in a 20 µL reaction containing 7.5 µL of deionized water, 10 µL of SYBR mix, 0.5 µL of forward primer (10 µM), 0.5 µL of reverse primer (10 µM), and 1.5 µL of template cDNA. The PCR procedure was as follows: 2 min of pre-denaturation at 95°C; 40 cycles of 20 s at 95°C, 25 s at 58°C, and 60 s at 68°C; and an additional step for dissociation (15 s at 95°C, 60 s at 60°C, and 15 s at 95°C). The 7000 System SDS Software (Applied Biosystems) was used for data collection. Relative transcript levels for each sample were determined using the comparative cycle threshold (Ct), and normalized against the geometric mean of the Ct value of three internal control genes (18s RNA; ubiquitin-conjugating enzyme-like protein, UBC; NAD-dependent glyceraldehyde-3-phosphate dehydrogenase, GAPDH). Expression ratios were calculated using the ΔΔCt method corrected for the PCR efficiency for each gene.
Results
Construction of the Masson pine transcript dataset
Even amounts of RNA collected from the seedlings grown under Pi-sufficient (control) or Pi-deficient (P1 and P2) conditions for 12, 24, 48 and 60 days were used to construct two cDNA libraries. Second-generation sequencing was then used to construct the transcript dataset. A total of 70,896 unigenes were assembled from 112,108,862 high quality RNA-seq reads generated from two libraries (as described in the Materials and methods). The length of unigenes ranged from 201 to 20,490 bp, and the total base count of these sequences was 70,169,217 bp (Figure 1).
Unigenes were analyzed for similarity/sequence conservation against non-redundant (nr) datasets of various species using BLAST (E-value ≤1e-5), and over 47,100 unigenes were identified in plants. Putative functions were assigned for 66.3% of the sequences. An additional >56,000 more highly expressed transcripts were manually curated to ensure accurate expression profile analysis of the microarray data.
Identification of transcripts differentially expressed under Pi-deficient conditions using microarray data
A series of microarray experiments were used to measure gene expression in Masson pine seedlings grown under conditions of low Pi for four periods. Using ANOVA with the Benjamini-Hochberg correction for multiple testing, we identified 1,396 and 943 differentially expressed genes (P<0.05) in the four periods under P1 and P2 conditions, respectively (Figure 2; Table S1, S2). Of the 1,396 transcripts identified in P1, 776 were up-regulated, while 544 of the 943 P2 transcripts followed the same trend. The number of transcripts induced in P2 was less than 32% of those induced in P1, suggesting that the growth of Masson pine was more severely limited by lower Pi levels. Analysis of overlap on the Venn diagram (Figure 2) showed that 179 transcripts were induced in P1 and P2. Of these, 14 encoded genes involved in transferring phosphorus-containing groups from donors to acceptors. Another 14 transcripts encoding genes similar to hydrolases that act on esters, glycosyl bonds and acid anhydrides, and a further eight genes encoded sequence-specific DNA binding transcription factors. Other identified genes included several pentatricopeptide repeat-containing proteins, gibberellin induced protein, thaumatin-like protein, and a further 66 cDNAs with unknown function (Figure 3; Table S3).
A total of 2,160 transcripts were identified as differentially expressed under both Pi-deficiency treatments.
Categorization of 179 transcripts based on molecular function.
Characterization of genes induced by Pi-deficiency
For functional annotation of transcripts up-regulated under low Pi conditions, Gene Ontology (GO) was used. The top 20 GO categories were identified for P1 and P2 stressed seedlings (Figure 4). Although the numbers of transcripts differed, the overall categorizations were similar between P1 and P2 stressed seedlings, with most involved in metabolic processes, including various transferases, phosphatases, protein kinases, synthases, hydrolases, polymerases and oxidases. Many glycosyltransferases and lipases that involved in lipid metabolism were included, which may alter the membrane lipid composition to maintain intracellular Pi homeostasis under Pi-deficiency [49]. Many P1 and P2 transcripts encoded redox proteins, especially cytochrome P450 enzymes and the 2OG-FeII oxidoreductase, both of which were identified as up-regulated in response to phosphate starvation in Arabidopsis and wheat [50], [51]. Other oxidation-reduction-associated protein genes included alcohol dehydrogenase, ascorbate oxidase (AO), laccase, polyphenol oxidase, primary amine oxidase, and flavonoid-associated enzymes were also differentially regulated as subjected to the Pi-deficiency. Most of the remaining differentially expressed transcripts belonged to the stress or stimulus category.
The numbers of up-regulated transcripts in P1 and P2 seedlings categorized in the top 20 GO groups are shown. The x-axis indicates the number of transcripts in a category, and the y-axis indicates the category.
The uptake of Pi from soil alters the expression of transporter proteins (TPs) and TFs, and the expression of these proteins is affected by the Pi abundance. Both TPs and TFs were differentially expressed in P1 and P2 seedlings, and there were differences between the two sets of treatments. A total of 20 different TPs were differentially expressed in P1, while only 10 were differentially expressed in P2 (Table S4). Five important phosphate TPs were only identified in P1 seedlings, and some potassium and drug transmembrane TPs were also only identified in P1 seedlings. Some TP transcripts identified in both P1 and P2 seedlings exhibited significantly different levels; five ABC transporters were highly differentially expressed in P1 while only three were up-regulated in P2, and six nitrate transporters were up-regulated in P1, while only one was present down-regulated in P2. Additionally, 14 TF family proteins were differentially expressed under Pi-deficiency stress conditions (Table S5). AP2, MYB, WRKY and MADS-Box proteins were identified as the major TFs responding to Pi-deficiency stress under both P1 and P2 conditions. Of these, all MADS-Box proteins were down-regulated, whereas, MYB and WRKY TFs were up-regulated. Five GATA TFs and one bHLH TF protein were down-regulated in P1 but not altered in P2.
As noted in previous studies, harmful reactive oxygen species (ROS) can be generated under abiotic stress conditions, including Pi-deficiency [16], [52]. The microarray data obtained in this study revealed up-regulation of glutathione s-transferase (GST, comp20392_c0_seq1) and glutathione peroxidase (comp28364_c0_seq1), and down-regulation of AO (comp25395_c1_seq1, comp13179_c0_seq1, comp12918_c0_seq1, comp12918_c2_seq1) under P1 condition. Additionally, SPX domain-containing genes were also up-regulated in P1 stressed seedlings.
Expression profiles of Pi-deficiency induced genes
To obtain insight into the mechanisms involved in the response to Pi-deficiency, transcriptomic analyses were performed on seedlings grown for different durations under the stress conditions. A self-organizing map (SOM) was used for clustering using the GeneSpring software. The expression profiles of differentially expressed transcripts under P1 and P2 treatments were divided into at least four groups (Figure 5), among which most transcripts were stably and gradually up- or down-regulated within 48 days. Interestingly, a number of transcripts increased or decreased much more rapidly and over a wider dynamic range under P1 conditions than did that under P2 conditions (Figure 5).
A time course of changes in expression pattern is shown. Seedlings under P0 conditions are included as a control. The x-axis represents time points after treatment, and the y-axis indicates the expression level.
Microarray expression validation by qRT-PCR
Generally, qRT-PCR is used to validate gene expression levels determined using high-throughput technologies such as microarray and RNA-seq. Therefore, we further confirmed the microarray expression profiles using qRT-PCR under identical conditions for a set of nine randomly selected unigenes involved in the phosphate starvation response. A high correlation was identified between qRT-PCR and microarray results, confirming the high reliability of the microarray expression data (Figure 6).
The x-axis indicates the duration of the Pi-deficiency stress, and the y-axis indicates the signal ratio. The conditions of P1 and P2 are described in the Materials and methods.
Discussion
Plant acclimation to Pi-deficiency stress is an important topic in plant biology, not least because of our dependence on Pi fertilizers, which are derived from a non-renewable natural resource that is being rapidly depleted. Several studies based on high-throughput technologies have highlighted the highly complex molecular systems involved in response to Pi-deficiency in plants, most of which focused on model plants or annual crops [16], [20], [50], [51]. In the present study, we focused on acclimation to Pi-deficiency in Masson pine using a transcriptomics approach based on a combination of RNA-seq and microarray experiments. We carefully designed the experiments to distinguish sets of genes that are involved in response to Pi-deficiency at different time points and at different levels of Pi-deficiency. A total of 1,396 differentially expressed transcripts were identified in P1 stress seedlings, and 943 were identified in P2 plants. A greater number of genes were therefore differentially expressed under conditions of lower Pi. Most of the differentially expressed genes fell into three major functional groups: metabolism, protein synthesis and stress response. To our knowledge, this is the first whole transcriptome analysis on Masson pine to address Pi-deficiency.
Expression profiles of genes involved in the response to Pi-deficiency
A slow-growing plant will need less nutrients over a particular time period than a fast-growing individual, since the demand for nutrients is correlated with dry weight production [53]. No visible anthocyanin accumulated in the seedlings used in the current study, and no other growth differences were apparent in the aerial tissues during the duration of the experiments, in agreement with previous observations [37]. Since Masson pine is a relatively slow-growing plant, seedlings from four time periods (see Materials and methods) were chosen to analyze gene expression differences. The results showed that most differentially expressed transcripts were stably up- or down-regulated, suggesting that some functional genes are expressed continuously and efficiently in slow-growing plants.
Alteration of lipid metabolism and membrane composition under Pi deficiency
The lipid composition of plant membranes can change drastically under conditions of Pi-deficiency, with an increase in non-phosphorus-containing lipids and a decrease in phospholipids [54]. Phosphate-deficient oat replaces a major portion of the plasma membrane phospholipids with the galactolipid digalactosyldiacylglycerol, and phospholipids are replaced by glycolipids in both the plasma membrane and tonoplast [55], [56]. Currently, an UDP-glycosytransferase (e.g., comp31295_c0_seq4, comp19255_c0_seq1, comp13888_c0_seq1) was up-regulated under P1 and P2 treatments, which indicated that the membrane lipid composition might have changed in response to low Pi stress.
Expression of genes involved in Pi sensing and transport
All transcripts identified were differentially expressed under P1 and P2 conditions. Of particular interest, transcription of genes involved in Pi sensing and transport exhibited a >2-fold change in expression levels under P1 conditions, whereas these genes were practically unaffected by the P2 treatment. SPX domain-containing proteins are involved in Pi signaling and transport, and are essential for maintaining Pi homeostasis [5], [57]–[60]. Twenty SPX domain-containing proteins have been identified in Arabidopsis. One of the transcripts identified in P1 treated seedlings is an orthologue of AtSPX-MFS1 (At4g22990), which is a predicted transmembrane protein targeted to the vacuole that is weakly induced by Pi-deficiency [59]. However, it is difficult to draw any functional comparisons between SPX isoforms in Masson pine and Arabidopsis from the currently available data. The SPX domain-containing gene (comp23145_c2_seq1) identified herein was markedly up-regulated in P1 treatment, but no correlation between expression levels and phenotypic traits were observed. A total of five phosphate transporters were also up-regulated in P1 seedlings, and expression of these genes does appear to correlate with expression of the SPX domain-containing protein. The transcriptional changes identified in the P1 seedlings displayed a much higher magnitude and dynamic range than those identified in the P2 treatments. In addition to phosphate TPs, a potassium transporter was also up-regulated in P1, suggesting that the transport of phosphorus might be co-regulated and/or coordinated with potassium in this species, as documented previously in tomato [61]. Pi-deficiency (both P1 and P2) also increased the expression patterns of a number of ABC transporters, which have been shown to be induced under various abiotic stress conditions [62]–[64]. An anion exchanger presumably involved in Pi removal and absorption is one such protein [65], [66]. Our results also indicated an increased transport of sugars (comp14752_c0_seq1, comp324303_c0_seq1), amino acids (comp31360_c0_seq5, comp20063_c0_seq1), and organic acids (comp16177_c0_seq1), any of which could be involved in Pi mobilization and acquisition.
Transcription factor expression was altered in response to Pi stress
The major TFs induced by Pi-deficiency (in both P1 and P2) belonged to the AP2, MYB, WRKY, and MADS-Box families. This is in contrast to previous studies from Arabidopsis in which the major TFs affected by Pi stress belonged to the PHR, WRKY, ZAT, and bHLH subfamilies. In Arabidopsis, WRKY75 was up-regulated by Pi-deficiency stress, and in turn positively regulated Pi starvation-inducible genes and negatively regulated root development [67]. WRKY6 and WRKY42 are involved in the response to Pi-deficiency by regulating PHOSPHATE1 (PHO1) expression [68], which plays a key role in Pi translocation from roots to shoots, aiding homeostasis [69]–[71]. In the present work, WRKY TFs were up-regulated in Masson pine under both P1 and P2 conditions (Table S5). The largest class of TFs induced by low Pi stress in our experiments was the MYB TF family. A total of 12 MYBs (10 in P1 and two in P2) were up-regulated by Pi-deficiency in Masson pine seedlings. One of these, Phosphate Starvation Response1 (PHR1), was previously identified as a regulator involved in Pi-deficiency that activates a subset of genes including SPX domain-containing proteins by binding to the P1BS element in their promoter regions [59], [72]. Another, AtMYB2, regulates the plant response to Pi starvation by regulating the expression of the miR399 gene in Arabidopsis [73]. Recently, Hernández et al. (2007) identified four TFs that were up-regulated in Pi-deficient common bean roots, three of which belonged to the MYB family [25]. Furthermore, a total of 33 MYBs were up-regulated in Pi deficient lupin leaves [16]. Therefore, MYB TF family proteins would be expected to play a role in the response to Pi-deficiency in Masson pine, which explains the results obtained. Plant MADS domain proteins were first identified as regulators of floral organ identity and had been found to control additional developmental processes such as flowering time, fruit dehiscence, meristem identity and root development [74]. Zhang and Forde (1998) identified a nitrate-inducible Arabidopsis gene (ANR1) that encodes a MADS-Box TF, and demonstrated that it controls nutrient-induced changes in root architecture [75]. Another MADS-Box gene (CaJOINTLESS) is involved in suppression of vegetative growth in shoot meristems in pepper [76]. Pi-deficiency can profoundly disturb root meristem maintenance [77], but it is still unknown whether MADS-Box genes are involved in acclimation to low Pi. In the current study, four MADS-Box transcripts were down-regulated in P1 seedlings and two were down-regulated in P2 treatment, suggesting that they may be involved in promoting meristem growth of Masson pine. The much higher root-to-shoot ratio (R/S) under Pi-deficiency conditions observed herein strongly indicated that down-regulation of MADS-Box TFs promoted root development in Masson pine to assist acclimation to low Pi conditions.
Responses to oxidative stress
Metabolic pathways continuously generate ROS as by-products that inflict oxidative damage to cellular components and serve as critical signaling molecules in cell proliferation and survival [16], [78]. ROS homeostasis is usually maintained via regulation of antioxidant levels, but biotic and abiotic stresses can disturb the equilibrium between ROS formation and the antioxidant response, resulting in oxidative stress. The magnitude of Pi-deficiency has been directly correlated with the severity of oxidative stress [79], [80]. In this study, microarray data showed that glutathione s-transferase and glutathione peroxidase, which are both involved in cellular detoxification [80], [81], were up-regulated in Pi-deficient conditions, in agreement with previous studies [16], [82]. AO is an enzyme localized at the cell wall that uses oxygen to catalyze the oxidation of ascorbate to generate the unstable monodehydroascorbate (MDHA) radical, which rapidly disproportionates to yield dehydroascorbate (DHA) and ascorbate, and thus contributes to regulation of the ascorbate redox state. AO-overexpressing transgenic plants were found to exhibit a reduced capacity to up-regulate plant defenses against ROS [83]. Furthermore, suppression of AO expression conferred resistance to oxidative stress [84]. It appears that negligible expression of AO is a biological marker of severe oxidative stress in plants. In the current case, AO was down-regulated in P1 seedlings, which provides further evidence that severe Pi-deficiency induces severe oxidative stress. In our previous investigation, superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) activities were up-regulated in Masson pine under the conditions of Pi-deficiency (unpublished data), which further indicates that Pi-deficiency disturbs ROS homeostasis.
Conclusion
To our knowledge, this work is the first report of a whole transcriptome analysis of Pi-deficiency in Masson pine, and firstly demonstrated the long-term transcriptomic changes that occur under different low-Pi growth conditions. A total of 1,396 transcripts were differentially expressed under severe Pi-deficiency (P1) and 943 transcripts were differentially expressed under moderate Pi-deficiency (P2). Stress-associated genes and transcription factors were particularly well represented, and some genes of unknown function were also up- or down-regulated. Most of the identified genes were consistent with previously reported studies on Pi-deficiency, but novel pathways were also identified. The available data suggested that the Pi-deficiency tolerance of Masson pine is a very complex physiological and biochemical process, in which many genes and multiple metabolism pathways were cumulatively implicated. This contribution was also significative to the development of genomic resources of Masson pine and other coniferous trees and should accelerate the progress of breeding programs and functional genomic studies.
Supporting Information
Table S1.
Transcripts differentially expressed during different durations of Pi-deficiency stress experienced by P1 seedlings. Values are normalized against the microarray data (P<0.05) and log 2 transformed. Where possible, annotations are included.
https://doi.org/10.1371/journal.pone.0105068.s001
(XLS)
Table S2.
Transcripts differentially expressed during different durations of Pi-deficiency stress experienced by P2 seedlings. Values are normalized against the microarray data (P<0.05) and log 2 transformed. Where possible, annotations are included.
https://doi.org/10.1371/journal.pone.0105068.s002
(XLS)
Table S3.
Transcripts differentially expressed in P1 and P2. 179 transcripts were calculated and normalized using GeneSpring GX. Where possible, annotations are included.
https://doi.org/10.1371/journal.pone.0105068.s003
(XLS)
Table S4.
Transporter proteins differentially expressed under Pi-deficiency stress.
https://doi.org/10.1371/journal.pone.0105068.s004
(XLS)
Table S5.
Transcription factors differentially expressed under Pi-deficiency stress.
https://doi.org/10.1371/journal.pone.0105068.s005
(XLS)
Acknowledgments
We would like to thank CapitalBio Corporation (China) for providing the sequencing services and microarray preparations.
Author Contributions
Conceived and designed the experiments: FF GD XW. Performed the experiments: FF BC TZ. Analyzed the data: FF GQ. Contributed reagents/materials/analysis tools: FF BC TZ GQ. Wrote the paper: FF XW.
References
- 1. Hinsinger P (2001) Bioavailability of soil inorganic P in the rhizosphere as affected by root-induced chemical changes: a review. Plant Soil 237(2): 173–195.
- 2. Shen J, Yan L, Zhang J, Li H, Bai Z, et al. (2011) Phosphorus dynamics: from soil to plant. Plant Physiol 156(3): 997–1005.
- 3. Vance CP, Uhde-Stone C, Allan DLP (2003) Phosphorus acquisition and use: critical adaptations by plants for securing a nonrenewable resource. New Phytol 157(3): 423–447.
- 4. Cordell D, Drangert J, White S (2009) The story of phosphorus: global food security and food for thought. Glob Environ Change 19(2): 292–305.
- 5. Chiou TJ, Lin SI (2011) Signaling network in sensing phosphate availability in plants. Annu Rev Plant Biol 62: 185–206.
- 6. Gaxiola RA, Edwards M, Elser JJ (2011) A transgenic approach to enhance phosphorus use efficiency in crops as part of a comprehensive strategy for sustainable agriculture. Chemosphere 84(6): 840–845.
- 7. Jain A, Nagarajan VK, Raghothama KG (2012) Transcriptional regulation of phosphate acquisition by higher plants. Cell Mol Life Sci 69(19): 3207–3224.
- 8. Ticconi CA, Abel S (2004) Short on phosphate: plant surveillance and countermeasures. Trends Plant Sci 9(11): 548–555.
- 9. Lynch JP (2011) Root phenes for enhanced soil exploration and phosphorus acquisition: tools for future crops. Plant Physiol 156(3): 1041–1049.
- 10. Wasaki J, Yamamura T, Shinano T, Osaki M (2003) Secreted acid phosphatase is expressed in cluster roots of lupin in response to phosphorus deficiency. Plant Soil 248(1–2): 129–136.
- 11. Baldwin JC, Karthikeyan AS, Cao A, Raghothama KG (2008) Biochemical and molecular analysis of the LePS2; 1: a phosphate starvation induced protein phosphatase gene from tomato. Planta 228(2): 273–280.
- 12. Carvalhais LC, Dennis PG, Fedoseyenko D, Hajirezaei MR, Borriss R, et al. (2011) Root exudation of sugars, amino acids, and organic acids by maize as affected by nitrogen, phosphorus, potassium, and iron deficiency. J Plant Nutr Soil Sci 174(1): 3–11.
- 13. Plaxton WC, Tran HT (2011) Metabolic adaptations of phosphate-starved plants. Plant Physiol 156(3): 1006–1015.
- 14. Raghothama KG, Karthikeyan AS (2005) Phosphate acquisition. Plant Soil 274: 37–49.
- 15. Loth-Pereda V, Orsini E, Courty PE, Lota F, Kohler A, et al. (2011) Structure and expression profile of the phosphate Pht1 transporter gene family in mycorrhizal Populus trichocarpa. Plant Physiol 156(4): 2141–2154.
- 16. O′Rourke JA, Yang SS, Miller SS, Bucciarelli B, Liu J, et al. (2013) An RNA-Seq transcriptome analysis of orthophosphate-deficient white lupin reveals novel insights into phosphorus acclimation in plants. Plant Physiol 161(2): 705–724.
- 17. Muchhal US, Pardo JM, Raghothama KG (1996) Phosphate transporters from the higher plant Arabidopsis thaliana. Proc Natl Acad Sci 93(19): 10519–10523.
- 18. Hammond JP, Bennett MJ, Bowen HC, Broadley MR, Eastwood DC (2003) Changes in gene expression in Arabidopsis shoots during phosphate starvation and the potential for developing smart plants. Plant Physiol 132(2): 578–596.
- 19. Bayle V, Arrighi JF, Creff A, Nespoulous C, Vialaret J, et al. (2011) Arabidopsis thaliana high-affinity phosphate transporters exhibit multiple levels of posttranslational regulation. Plant Cell 23(4): 1523–1535.
- 20. Woo J, MacPherson CR, Liu J, Wang H, Kiba T, et al. (2012) The response and recovery of the Arabidopsis thaliana transcriptome to phosphate starvation. BMC Plant Biol 12(1): 62.
- 21. Wasaki J, Shinano T, Onishi K, Yonetani R, Yazaki J, et al. (2006) Transcriptomic analysis indicates putative metabolic changes caused by manipulation of phosphorus availability in rice leaves. J Exp Bot 57(9): 2049–2059.
- 22. Calderon-Vazquez C, Ibarra-Laclette E, Caballero-Perez J, Herrera-Estrella L (2008) Transcript profiling of Zea mays roots reveals gene response to phosphate deficiency at the plant- and species-specific levels. J Exp Bot 59(9): 2479–2497.
- 23. Hammond JP, Mayes S, Bowen HC, Graham NS, Hayden RM, et al. (2011) Regulatory hotspots are associated with plant gene expression under varying soil phosphorus supply in Brassica rapa. Plant Physiol 156(3): 1230–1241.
- 24. Fita A, Bowen HC, Hayden RM, Nuez F, Picó B, et al. (2012) Diversity in expression of phosphorus (P) responsive genes in Cucumis melo L. PLoS One. 7(4): e35387.
- 25. Hernández G, Ramírez M, Valdés-López O, Tesfaye M, Graham MA, et al. (2007) Phosphorus stress in common bean: Root transcript and metabolic responses. Plant Physiol 144(2): 752–767.
- 26. Fang ZY, Shao C, Meng YJ, Wu P, Chen M (2009) Phosphate signaling in Arabidopsis and Oryza sativa. Plant Sci 176(2): 170–180.
- 27. Duff SMG, Sarath G, Plaxton WC (1994) The role of acid-phosphatases in plant phosphorus-metabolism. Physiol Plantarum 90(4): 791–800.
- 28. Raghothama KG (2000) Phosphate transport and signaling. Curr Opin Plant Biol 3(3): 182–187.
- 29. Lan P, Li W, Schmidt W (2012) Complementary proteome and transcriptome profiling in phosphate-deficient Arabidopsis roots reveals multiple levels of gene regulation. Mol Cell Proteomics 11(11): 1156–1166.
- 30. Yang XJ, Finnegan PM (2010) Regulation of phosphate starvation responses in higher plants. Ann Bot (Lond) 105(4): 513–526.
- 31. Gamuyao R, Chin JH, Pariasca-Tanaka J, Pesaresi P, Catausan S, et al. (2012) The protein kinase Pstol1 from traditional rice confers tolerance of phosphorus deficiency. Nature 488(7412): 535–539.
- 32. Hackenberg M, Huang PJ, Huang CY, Shi BJ, Gustafson P, et al. (2013) A comprehensive expression profile of microRNAs and other classed of non-coding small RNAs in barley under phosphorous-deficient and sufficient conditions. DNA Res 20(2): 109–125.
- 33. Zhang Y, Zhou ZC, Ma XH, Jin G (2010) Foraging ability and growth performance of four subtropical tree species in response to heterogeneous nutrient environments. J For Res 15(2): 91–98.
- 34. Zhou ZC, Xie YR, Jin GQ, Chen Y, Song ZY (2005) Study on phosphorus efficiency of different provenances of Pinus massoniana. Scientia Silvae Sinicae 41(4): 25–30 (in Chinese)..
- 35. Xie YR, Zhou ZC, Jin GQ, Chen Y, Song ZY (2004) Root morphology and dry matter allocation of masson pine: response of different provenances to low phosphorus stress. Forest Research 17(3): 278–282.
- 36. Xie YR, Zhou ZC, Liao GH, Jin GQ, Chen Y (2005) Difference of induced acid phosphate activity under low phosphorus stress of Pinus massoniana provenances. Scientia Silvae Sinicae 41(3): 58–62 (in Chinese)..
- 37. Zhang Y, Zhou ZC, Yang Q (2013) Genetic variations in root morphology and phosphorus efficiency of Pinus massoniana under heterogeneous and homogeneous low phosphorus conditions. Plant Soil 364(1–2): 93–104.
- 38. Malone JH, Oliver B (2011) Microarrays, deep sequencing and the true measure of the transcriptome. BMC Biology 9(1): 34.
- 39. AC′t Hoen PA, Ariyurek Y, Thygesen HH, Vreugdenhil E, Vossen RH, et al. (2008) Deep sequencing-based expression analysis shows major advances in robustness, resolution and inter-lab portability over five microarray platforms. Nucleic Acids Res 36(21): e141.
- 40. Marioni JC, Mason CE, Mane SM, Stephens M, Gilad Y (2008) RNA-seq: an assessment of technical reproducibility and comparison with gene expression arrays. Genome Res 18(9): 1509–1517.
- 41. Fu X, Fu N, Guo S, Yan Z, Xu Y, et al. (2009) Estimating accuracy of RNA-seq and microarrays with proteomics. BMC Genomics 10(1): 161.
- 42. Su Z, Li Z, Chen T, Li QZ, Fang H, et al. (2011) Comparing next-generation sequencing and microarray technologies in a toxicological study of the effects of aristolochic acid on rat kidneys. Chen Res Toxicol 24(9): 1486–1493.
- 43. Kogenaru S, Yan Q, Guo Y, Wang N (2012) RNA-seq and microarray complement each other in transcriptome profiling. BMC Genomics 13(1): 629.
- 44. Tatusov RL, Koonin EV, Lipman DJ (1997) A genomic perspective on protein families. Science 278(5338): 631–637.
- 45. Tatusov RL, Fedorova ND, Jackson JD, Jacobs AR, Kiryutin B, et al. (2003) The COG database: an updated version includes eukaryotes. BMC Bioinformatics 4(1): 41.
- 46. Mulder NJ, Apweiler R, Attwood TK, Bairoch A, Barrell D, et al. (2003) The InterPro Database, 2003 brings increased coverage and new features. Nucleic Acids Res 31(1): 315–318.
- 47. Zdobnov EM, Apweiler R (2001) InterProScan an integration platform for the signature-recognition methods in InterPro. Bioinformatics 17(9): 847–848.
- 48. Harris MA, Clark J, Ireland A, Lomax J, Ashburner M, et al. (2004) The Gene Ontology (GO) database and informatics resource. Nucleic Acids Res 32: D258–61.
- 49. Liu WY, Lin SI, Chiou TJ (2009) Molecular regulators of phosphate homeostasis in plants. J Exp Bot 60(5): 1427–1438.
- 50. Thibaud MC, Arrighi JF, Bayle V, Chiarenza S, Creff A, et al. (2010) Dissection of local and systemic transcriptional responses to phosphate starvation in Arabidopsis. Plant J 64(5): 775–789.
- 51. Oono Y, Kobayashi F, Kawahara Y, Yazawa T, Handa H, et al. (2013) Characterisation of the wheat (triticum aestivum L.) transcriptome by de novo assembly for the discovery of phosphate starvation-responsive genes: gene expression in Pi-stressed wheat. BMC Genomics 14(1): 77.
- 52. Misson J, Raghothama KG, Jain A, Jouhet J, Block MA, et al. (2005) A genome-wide transcriptional analysis using Arabidopsis thaliana Affymetrix gene chips determined plant responses to phosphate deprivation. Proc Natl Acad Sci 102(33): 11934–11939.
- 53. Krouk G, Ruffel S, Gutiérrez RA, Gojon A, Crawford NM, et al. (2011) A framework integrating plant growth with hormones and nutrients. Trends Plant Sci 16(4): 178–182.
- 54. Russo MA, Quartacci MF, Izzo R, Belligno A, Navari-Izzo F (2007) Long-and short-term phosphate deprivation in bean roots: plasma membrane lipid alterations and transient stimultion of phospholipases. Phytochemistry 68(11): 1564–1571.
- 55. Andersson MX, Stridh MH, Larsson KE, Liljenberg C, Sandelius AS (2003) Phosphate-deficient oat replaces a major portion of the plasma membrane phospholipids with the galactolipid digalactosyldiacylglycerol. FEBS Lett 537(1–3): 128–132.
- 56. Andersson MX Larsson KE, Tjellström H, Liljenberg C, Sandelius AS (2005) Phosphate-limited oat The plasma membrane and the tonoplast as major targets for phospholipids- to- glycolipid replacement and stimulation of phospholipases in the plasma membrane. J Biol Chem 280(30): 27578–27586.
- 57. Duan K, Yi K, Dang L, Huang H, Wu W, et al. (2008) Characterization of a sub-family of Arabidopsis genes with the SPX domain reveals their diverse functions in plant tolerance to phosphorus starvation. Plant J 54(6): 965–975.
- 58. Yang G, Ding G, Shi L, Cai H, Xu F (2012) Characterization of phosphorus starvation-induced gene BnSPX3 in Brassica napus. Plant Soil 350(1–2): 339–351.
- 59. Secco D, Wang C, Arpat BA, Wang Z, Poirier Y, et al. (2012) The emerging importance of the SPX domain-containing proteins in phosphate homeostasis. New Phytol 193(4): 842–851.
- 60. Wang C, Huang W, Ying Y, Li S, Secco D, et al. (2012) Functional characterization of the rice SPX-MFS family reveals a key role of OsSPX- MFS1 in controlling phosphate homeostasis in leaves. New Phytol 196(1): 139–148.
- 61. Wang YH, Garvin DF, Kochian LV (2002) Rapid induction of regulatory and transport genes in response to phosphorus, potassium, and iron deficiencies in tomato roots: evidence for cross talk and root/rhizosphere-mediated signals. Plant Physiol 130(3): 1361–1370.
- 62. Yazaki K (2006) ABC transporters involved in the transport of plant secondary metabolites. FEBS lett 580(4): 1183–1191.
- 63. Huang CF, Yamaji N, Ma JF (2010) Knockout of a bacterial-type ATP-binding cassette transporter gene, AtSTAR1, results in increased aluminum sensitivity in Arabidopsis. Plant Physiol 153(4): 1669–1677.
- 64. Xi J, Xu P, Xiang CB (2012) Loss of AtPDR11, a plasma membrane-localized ABC transporter, confers paraquat tolerance in Arabidopsis thaliana. Plant J 69(5): 782–791.
- 65. Xu X, Gao BY, Yue QY, Zhong QQ (2010) Preparation of agricultural by-product based anion exchanger and its utilization for nitrate and phosphate removal. Bioresource Technol 101(22): 8558–8564.
- 66. Awual MR, Jyo A, EI-Safty SA, Tamada M, Seko N (2011) A weak-base fibrous anion exchanger effective for rapid phosphate removal from water. J Hazard Mater 188(1): 164–171.
- 67. Devaiah BN, Karthikeyan AS, Raghothama KG (2007) WRKY75 transcription factor is a modulator of phosphate acquisition and root development in Arabidopsis. Plant Physiol 143(4): 1789–1801.
- 68. Chen YF, Li LQ, Xu Q, Kong YH, Wang H, et al. (2009) The WRKY6 transcription factor modulates PHOSPHATE1 expression in response to low Pi stress in Arabidopsis. Plant Cell 21(11): 3554–3566.
- 69. Hamburger D, Rezzonico E, MacDonald-Comber Petetot J, Somerville C, Poirier Y (2002) Identification and characterization of the Arabidopsis PHO1 gene involved in phosphate loading to the xylem. Plant Cell 14(4): 889–902.
- 70. Secco D, Baumann A, Poirier Y (2010) Characterization of the rice PHO1 gene family reveals a key role for OsPHO1; 2 in phosphate homeostasis and the evolution of a distinct clade in dicotyledons. Plant Physiol 152(3): 1693–1704.
- 71. Liu TY, Huang TK, Tseng CY, Lai YS, Lin SI, et al. (2012) PHO2-dependent degradation of PHO1 modulates phosphate homeostasis in Arabidopsis. Plant Cell 24(5): 2168–2183.
- 72. Nilsson L, Müller R, Nielsen TH (2007) Increased expression of the MYB-related transcription factor, PHR1, leads to enhanced phosphate uptake in Arabidopsis thaliana. Plant Cell Environ 30(12): 1499–1512.
- 73. Baek D, Kim MC, Chun HJ, Kang S, Park HC, et al. (2013) Regulation of miR399f transcription by AtMYB2 affects phosphate starvation responses in Arabidopsis. Plant Physiol 161(1): 362–373.
- 74. Riechmann JL, Heard J, Martin G, Reuber L, Jiang CZ, et al. (2000) Arabidopsis transcription factors: genome-wide comparative analysis among Eukaryotes. Science 290(5499): 2105–2110.
- 75. Zhang H, Forde BG (1998) An Arabidopsis MADS box gene that controls nutrient-induced changes in root architecture. Science 279(5349): 407–409.
- 76. Cohen O, Borovsky Y, David-Schwartz R, Paran I (2012) CaJOINTLESS is a MADS-box gene involved in suppression of vegetative growth in all shoot meristems in pepper. J Exp Bot 63(13): 4947–4957.
- 77. Desnos T (2008) Root branching responses to phosphate and nitrate. Curr Opin Plant Biol 11(1): 82–87.
- 78. Ray PD, Huang BW, Tsuji Y (2012) Reactive oxygen species (ROS) homeostasis and redox regulation in cellular signaling. Cell Signal 24(5): 981–990.
- 79. Shin R, Berg RH, Schachtman DP (2005) Reactive oxygen species and root hairs in Arabidopsis root response to nitrogen, phosphorus and potassium deficiency. Plant Cell Physiol 46(8): 1350–1357.
- 80. Torabi S, Wissuwa M, Heidari M, Naghavi MR, Gilany K, et al. (2009) A comparative proteome approach to decipher the mechanism of rice adaptation to phosphorous deficiency. Proteomics 9(1): 159–170.
- 81. Bhabak KP, Mugesh G (2010) Functional mimics of glutathione peroxidase: bioinspired synthetic antioxidants. Accounts Chem Res 43(11): 1408–1419.
- 82. Yao Y, Sun H, Xu F, Zhang X, Liu S (2011) Comparative proteome analysis of metabolic changes by low phosphorus stress in two Brassica napus genotypes. Planta 233(3): 523–537.
- 83. Fotopoulos V, Sanmartin M, Kanellis AK (2006) Effect of ascorbate oxidase over-expression on ascorbate recycling gene expression in response to agents imposing oxidative stress. J Exp Bot 57(14): 3933–3943.
- 84. Yamamoto A, Bhuiyan MNH, Waditee R, Tanaka Y, Esaka M, et al. (2005) Suppressed expression of the apoplastic ascorbate oxidase gene increases salt tolerance in tobacco and Arabidopsis plants. J Exp Bot 56(417): 1785–1796.