Figures
Abstract
Salinity stress has significant negative effects on plant biomass production and crop yield. Salinity tolerance is controlled by complex systems of gene expression and ion transport. The relationship between specific features of mild salinity stress adaptation and gene expression was analyzed using four commercial varieties of bread wheat (Triticum aestivum) that have different levels of salinity tolerance. The high-throughput phenotyping system in The Plant Accelerator at the Australian Plant Phenomics Facility revealed variation in shoot relative growth rate and salinity tolerance among the four cultivars. Comparative analysis of gene expression in the leaf sheaths identified genes whose functions are potentially linked to shoot biomass development and salinity tolerance. Early responses to mild salinity stress through changes in gene expression have an influence on the acquisition of stress tolerance and improvement in biomass accumulation during the early “osmotic” phase of salinity stress. In addition, results revealed transcript profiles for the wheat cultivars that were different from those of usual stress-inducible genes, but were related to those of plant growth. These findings suggest that, in the process of breeding, selection of specific traits with various salinity stress-inducible genes in commercial bread wheat has led to adaptation to mild salinity conditions.
Citation: Takahashi F, Tilbrook J, Trittermann C, Berger B, Roy SJ, Seki M, et al. (2015) Comparison of Leaf Sheath Transcriptome Profiles with Physiological Traits of Bread Wheat Cultivars under Salinity Stress. PLoS ONE 10(8): e0133322. https://doi.org/10.1371/journal.pone.0133322
Editor: Guangxiao Yang, Huazhong University of Science & Technology(HUST), CHINA
Received: March 28, 2015; Accepted: June 24, 2015; Published: August 5, 2015
Copyright: © 2015 Takahashi 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
Data Availability: All microarray data files are available from the ArrayExpress database (https://www.ebi.ac.uk/arrayexpress/; accession number E-MTAB-3270).
Funding: The Japan Society for the Promotion of Science special travel grant award for young scientists (http://www.jsps.go.jp/english/index.html), the Programme for Promotion of Basic and Applied Researches for Innovations in Bio-oriented Industry (http://www.naro.affrc.go.jp/english/index.html), the Ministry of Agriculture, Forestry and Fisheries (http://www.maff.go.jp/e/index.html), the RIKEN Biomass Engineering Program Division (http://www.riken.jp/bmep/english/index.html), the grants from the Australian and South Australian governments (http://www.australia.gov.au/), the grants from the Australian Research Council (http://www.arc.gov.au/), the Grains Research Development Corporation (http://www.grdc.com.au/), and King Abdullah University of Science and Technology (http://www.kaust.edu.sa/) had important role in study design, data collection and analysis, decision to publish, and preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Bread wheat (Triticum aestivum) is now grown on a greater area than any other cereal crop, but is only moderately tolerant to salinity stress [1]. With soil salinity affecting one-fifth of irrigated agricultural land worldwide [2], there is an urgent need to improve the salinity tolerance of bread wheat to meet future global food requirements.
Shoot Na+ exclusion is only one component of plant salinity tolerance, and some studies have found no clear correlation between plant salinity tolerance and Na+ exclusion under conditions of moderate salinity stress [3–6]. Another important, but not well characterized, tolerance trait is the shoot ion accumulation independent tolerance mechanism (also called “osmotic tolerance”), which is the ability of plants to maintain growth rates after initial exposure to salinity stress that is independent of the extent of shoot ion accumulation [1, 7]. This osmotic tolerance mechanism is important for maintaining tissue expansion and tillering during the early stages of salinity stress before Na+ and Cl- ions accumulate to toxic concentrations in the shoot. Plants with poor osmotic tolerance will show significant reduction in growth immediately after salt application. This reduction in growth will continue throughout the time of exposure to salinity, resulting in plants with reduced biomass and yield [1, 5, 7]. Unfortunately, the mechanisms behind osmotic tolerance are poorly understood, and no gene for increased tolerance has yet been linked specifically to this trait. To advance breeding for salinity tolerance in bread wheat, it is imperative to identify the genes important for osmotic stress tolerance.
It is well known that thousands of genes are differentially regulated in response to a range of abiotic stresses including salinity and drought [8, 9], due in part to plant stress-tolerance mechanisms in the early phase of stress responses. In wheat, several genes have been reported to be induced rapidly after salinity stress. The salinity stress-inducible transcription factor (TF), TaWRKY10, has been found to be important in regulating drought responses and the accumulation of reactive oxygen species [10]. Wheat calcineurin B-like protein-interacting protein kinase 29 (TaCIPK29) and sucrose nonfermenting 1-related protein kinase 2 (TaSnRK2) are also induced by abiotic stress and are important in plant stress signaling pathways that allow rapid control of multiple tolerance mechanisms [11–18]. However, it is still unclear whether these salinity stress-inducible genes are associated with salt or osmotic stress tolerance.
In the present study, four commercial bread wheat cultivars were exposed to salinity stress conditions to characterize the transcriptional response to early stages of osmotic stress in the key tissue affecting growth over this time period, the leaf sheaths. The high-throughput phenotyping system of The Plant Accelerator at the Australian Plant Phenomics Facility in South Australia allowed the continuous measurement of growth rates to determine the osmotic stress tolerance component of all four cultivars. A comprehensive evaluation based on gene expression profiles of leaf sheaths and phenomics highlighted gene sets associated with individual physiological traits. We discuss the correlation between the salinity stress tolerance of these commercial salinity-tolerant cultivars and their leaf sheath transcriptomes during the early response of shoots to salinity stress.
Materials and Methods
Plant material and growth conditions
Four commercial varieties of bread wheat (Triticum aestivum) were used in this study–Berkut, Krichauff, Gladius and Drysdale. Gladius and Drysdale are parents of a mapping population developed in Adelaide, and Berkut and Krichauff are the parents of another mapping population. Krichauff (pedigree: Wariquam//Kloka/Pitic62/3/Warimek/Halberd/4/3Ag3Aroona), Gladius and Drysdale (pedigree: Hartog*3/Quarrion) are relatively recent commercial Australian cultivars; Berkut (pedigree: Irena/Babax//Pastor) was developed by the International Maize and Wheat Improvement Centre (CIMMYT) in Mexico. Gladius was derived from a series of crossings among derivatives of RAC875, Krichauff, Excalibur, and Kukri. No specific permissions were required for the cultivation in the greenhouse of The Plant Accelerator at the Australian Plant Phenomics Facility described in the text because the non-transformant crops were grown and analyzed in the hermetically-closed greenhouse approved for academic plant science. The field studied did not involve endangered or protected species.
Seeds were soaked in reverse osmosis (RO) water at room temperature for 2 h. The water was drained and the seeds placed in the dark at 4°C for 2 days in sealed 50 mL polypropylene tubes. Evenly sized imbibed seeds were selected for sowing and four seeds per pot were planted 2 cm deep into 145 cm diameter by 190 cm height free-draining pots. The soil substrate was a potting mix consisting of 50% (v/v) University of California mix, 35% (v/v) peat mix and 15% (v/v) clay loam, designed specifically for use on The Plant Accelerator’s conveyor system. Plants were watered to a gravimetric water content of 25% (w/w). The seedlings were grown in natural light in a temperature-controlled greenhouse, with a daytime temperature of 22°C and 15°C at night. The pots were randomly placed in the greenhouse. At the time of emergence of leaf 2, seedlings were thinned, leaving one seedling per pot. Two experiments were performed with different concentrations of NaCl to measure the effects of mild salinity stress. The first experiment measured the growth response of wheat seedlings treated with 0 and 75 mM NaCl. The second experiment measured the growth response of seedlings treated with 0 and 100 mM NaCl, with RNA also extracted from leaf sheath tissue for microarray analysis. The experiments were conducted consecutively during the winter months of 2011 (from late May until early August) in Adelaide, South Australia (34°58'17.00"S, 138°38'23.00"E).
Assay for growth under salt stress conditions
Following emergence of the third leaf, the free-draining pots were placed in a deep saucer and loaded onto the conveyers of The Plant Accelerator’s automated imaging system [19], using a randomized block design (Brien et al, 2013), as shown in S1 and S2 Figs. Daily imaging was performed as described by Brien et al. (2013). Seedlings were watered to 25% (w/w) gravimetric water content. At the emergence of leaf four, the salinity treatment was applied as described by Berger et al. (2012). The salinity treatment solution (in a volume of 210 mL) was added to the saucer for uptake through the base of the pot, a process requiring approximately 2 h. After treatment, the soil water content was measured to be in the order of 35% (w/w).
For the first experiment, after application of the treatments, the pots contained 0 or approximately 50 mM NaCl which, over a period of 6 to 7 days was maintained at 0 or increased to 75 mM by evaporation and transpiration to reach the target gravimetric water content of 25% (w/w). For the second experiment, after treatment the soil contained 0 and approximately 70 mM NaCl and was maintained at 0 or increased to 100 mM NaCl over 6 to 7 days, as described above. Watering recommenced when the soil water content reached 25% (w/w) and NaCl in the soil solution was at the designated concentration. The method was designed to impose gradual salinity stress on the seedlings.
The LemnaTec Scanalyzer 3D (LemnaTec GmbH, Aachen, Germany) at The Plant Accelerator was used for daily image acquisition [20]. Two red-green-blue (RGB) images were captured from the side at a 90° rotation to each other and one RGB image from above. LemnaGrid software (LemnaTec GmbH, Aachen, Germany) was used for image processing to determine the projected shoot area—the sum of pixels identified as being part of the plant in each image [19, 21].
To measure shoot growth during the plant osmotic response period, the projected shoot area from days 0 to 8 after salinity application was recorded. At this stage, plants were still young and the use of an exponential growth model is still appropriate. There is also a linear correlation between projected shoot area and shoot biomass, for Triticum monococcum [22] and for bread wheat (B. Berger, unpublished results), so that projected shoot area can be used directly to measure relative growth rates (RGRs). The projected shoot area was plotted over time and exponential growth curves were fitted to determine RGR [19, 22]. The salinity tolerance index (STI) was determined:
Over the time course of this experiment, this approximates to a measure of the tolerance of plants to the osmotic phase of salinity toxicity, the shoot ion accumulation independent tolerance (Roy et al., 2014).
Stress treatment and extraction of RNA for microarrays
Separate replicate pots of each cultivar were grown under conditions identical to those of the second experiment (0 and 100 mM NaCl) for destructive sampling of shoot material for RNA extraction at four separate time points. The leaf sheaths of seedlings of each cultivar were collected immediately before (day 0) or 1, 2, and 3 days after application of 0 or 100 mM NaCl. Samples from three replicate seedlings were taken at each time point under both control and salt stress conditions, and total RNA from the sheaths was isolated using Trizol-modified reagent [23], and then used for the preparation of Cyanin-3 (Cy3)-labelled cRNA probes. Dye incorporation and cRNA yield were checked with the NanoDrop ND-1000 Spectrometer. The Cy3-labelled cRNA was fragmented and hybridized to the Agilent wheat microarray (Design ID: 0222297) according to the manufacturer’s guidelines. All microarray data have been deposited in the ArrayExpress database (https://www.ebi.ac.uk/arrayexpress/; accession number E-MTAB-3270).
Data mining of gene expression with Agilent wheat gene expression microarray
Microarray experiments were performed with the four bread wheat cultivars, Berkut, Krichauff, Gladius and Drysdale, using a 44K wheat gene expression array with a one-color gene expression microarray solution (Agilent Technologies). Each analysis was repeated in triplicate as biological replicates—a total of 84 microarray experiments were performed according to the manufacturer’s manual. Genes displaying a signal value of >1000 were selected for analysis. Feature extraction and image analysis software (GeneSpring GX version 12.5; Agilent Technologies) was used to locate and delineate each spot in the array and to perform a statistical analysis of the intensity of each spot using the Lowess method (p < 0.01). The statistical significance of gene expression was tested using the ANOVA analysis of variance test combined with a Benjamini and Hochberg false-discovery-rate multiple correction algorithm with a corrected q value of <0.01. The criterion for the fold change threshold selection was a greater than five-fold change of expression. The changes in gene expression were calculated as follows:
For expression analysis under control conditions:
Or, for expression analysis under salinity conditions:
To identify functional descriptions of genes, EST or probe sequence data were searched against the following nucleotide datasets using the EST data of the NCBI (http://www.ncbi.nlm.nih.gov/), the cDNA data of rice derived from RAP-DB v.2 (http://rapdb.dna.affrc.go.jp/) and the cDNA data present in TAIR release 10 (ftp://ftp.arabidopsis.org/home/tair/Sequences/blast_dataset/). The top-scoring hit for each query was selected for gene annotation with a threshold E-value of <1E-5, and the Load ID was used for the functional classification using the gene ontology (GO) of agriGO (http://bioinfo.cau.edu.cn/agriGO/), amiGO 2 (http://amigo.geneontology.org/amigo) and the Gene Ontology (http://www.geneontology.org/). The percentage of identified genes was calculated as follows:
Gene expression profile charts under stress conditions were searched against the following datasets using the eFP Browser (http://bbc.botany.utoronto.ca/efp/cgi-bin/efpWeb.cgi) for Arabidopsis and the RiceXPro (http://ricexpro.dna.affrc.go.jp/) for rice.
Quantitative RT-PCR analysis
Five microgram of total RNA was used for cDNA synthesis with random hexamer primers and SuperScript III reverse transcriptase (Invitrogen). SYBR Premis Ex Taq (TaKaRa) and gene-specific primers provided for the Primer Express Software v3.0.1 (Life technologies) were used for the reactions. Quantitative RT-PCR (qRT-PCR) analyses were performed using 7500 Fast Real-Time PCR systems (Applied Biosystems). The relative expression levels in each transcript were obtained by normalization to the actin gene (AB181991). The gene-specific primer sets were listed in S1 Table.
Results
Continuous monitoring of shoot biomass production reveals differences in growth responses to salinity between the bread wheat cultivars
Four commercial bread wheat cultivars were investigated for variation in salinity tolerance: Berkut, Krichauff, Gladius and Drysdale. Previous studies indicate that the doubled-haploid (DH) mapping population from a cross between Berkut and Krichauff were analyzed as salinity tolerance (ST) quantitative trait loci (QTL) to improve the grain yield-independent ST in field [24, 25]. A high-throughput phenotyping system was used to measure the growth of these bread wheat cultivars, and to quantify the effects of salinity on their growth. To evaluate the shoot ion independent component of salinity stress, the relative growth rate (RGR) of plants was estimated using the images of the projected shoot area for 8 days after treatment, and exponential growth curves fitted through the projected shoot area data [21, 26, 27]. For the relatively short time period of this experiment, the assumption of exponential growth was found to be a satisfactory approximation (Fig 1 and S3 Fig).
Quantification of growth through time under conditions of salinity stress of (A) Berkut (blue diamonds, SE, n = 6), (B) Krichauff (orange triangles, SE, n = 6,), (C) Gladius (green squares, SE, n = 4) or (D) Drysdale (red circles, SE, n = 4). Cultivars were treated with RO water (open symbols) or 100 mM NaCl (filled symbols) at the emergence of the fourth leaf (vertical line indicates day 0 or day of treatment). Projected shoot area (pixels) was determined by image analysis from digital images taken with RGB cameras. Exponential curves were fitted to data from day 0 to day 7 after treatment (rectangle within graphs A to D) to calculate relative growth rates for (E) Berkut, (F) Krichauff, (G) Gladius and (H) Drysdale seedlings under control and saline conditions.
Over the time course of the experiments, Berkut showed the highest mean RGR under control conditions (Figs 1 and 2). Treatment with 75 mM and 100 mM NaCl caused little reduction in RGR for both Drysdale and Krichauff, and a greater reduction in RGR under salinity was observed for Gladius and Berkut (Fig 2 and S3E Fig). These differences can be best summarized by calculating the salinity tolerance index (STI: Table 1), which indicated that Krichauff had the highest salinity tolerance under those salinity stress conditions (strictly, the greatest ability to maintain RGR over this time period in saline conditions relative to non-saline conditions).
(SE, n = 6 for Berkut and Krichauff, n = 4 for Gladius and Drysdale). RGRs of treated plants were significantly difference to RGRs of control plants (2 way ANOVA, p = 0.025).
Cluster analysis of variation in gene expression in leaf sheaths with respect to shoot growth and salinity stress tolerance
Microarray profiling of gene transcripts from leaf sheaths was performed to identify genes responding to the early stages of salinity stress. We selected the treatment of wheat plants with 100 mM NaCl for microarray analysis because the treatment with 100 mM NaCl affected the further widening of the difference of STI in comparison with the treatment of wheat with 75 mM NaCl among four cultivars (Table 1). For each cultivar, analysis was performed for both control and 100 mM NaCl treated plants on days 1, 2, and 3 after treatment. Integrated gene expression profiles for the treatment periods indicated hierarchical clustering side-by-side with control and moderate salinity stress conditions in each cultivar (S4A Fig). A 3D principal components analysis (PCA) revealed that the gene expression clusters for these cultivars differ significantly from each other even under control conditions (S4B Fig). Moreover, the large differences in expression of genes were observed between cultivars in samples from salt stressed plants, indicating that the gene expression profiles are different among different cultivars under both control and salinity stress conditions.
Although the patterns of leaf sheath transcriptome regulation varied sharply among the cultivars, an attempt was made to identify common responsive genes governing shoot biomass development and salinity tolerance. We focused first on RGRs under control conditions. Berkut showed the highest mean RGR under control conditions in both experiments (Fig 2 and S3 Fig). Comparative gene expression analysis revealed that 39 genes were upregulated only in Berkut during the first 3 days of growth under control conditions compared with the starting point (day 0) of the experiments (Fig 3A).
(A) Venn diagrams showing a five-fold or greater difference in expression with all possible regressions during the first 3 days of growth under control conditions compared with the starting point (day 0) of the experiments. (B) Heat maps indicating intensity of gene expression in Berkut. Log FC = log2 (the signal intensity under saline conditions / the signal intensity under control conditions). (C) Analysis of the gene ontology of 39 genes upregulated only in Berkut under control conditions. Functional categorizations by annotation were shown as gene ontology of biological process.
The timing of expression of the 39 genes exhibited three tendencies. The first, shown as cluster I, comprises the early response that occurred within 24 h after control conditions (Fig 3B). Cluster II comprises expressed genes in the second day under control conditions. Cluster III includes the late response that occurred on the 3rd day. All 39 genes were searched against the annotated genome datasets of both Arabidopsis and rice (TAIR 10 and RAP-DB) to identify their function. A comparative gene annotation analysis of wheat with rice and Arabidopsis indicated that most homologous genes have known functions (Table 2). Functional category classification using the gene ontology (GO) term revealed that most of the 39 genes were assigned to cellular and metabolic processes, developmental processes, and cell organization and biogenesis, suggesting that Berkut has a specific ability to maintain a high growth rate in the early stages of growth by transcription of particular genes involved in plant growth (Fig 3C).
Drysdale and Krichauff showed higher STI with the 100 mM NaCl treatment in comparison with Gladius and Berkut (Table 1), and they showed specific temporal variations in gene expression in response to salinity stress (Fig 4). Most upregulated and downregulated genes were expressed within 24 h of salinity treatment in both cultivars (Fig 4A). Of the 47 genes in Drysdale and 96 genes in Krichauff that were upregulated under salinity stress compared with control (RO water) conditions, 42 (89.4%) and 91 (94.8%), respectively, were observed to be upregulated one day after treatment. However, only five genes were similarly regulated in both cultivars (Fig 5A and Table 3). In contrast, the more salt sensitive Gladius and Berkut varieties did not exhibit dramatic changes in the numbers of upregulated genes during the three days after stress treatment, with only about 10 genes upregulated by mild salinity stress in each of the varieties. The same tendency was observed in the gene sets which were downregulated under mild salinity stress conditions (Figs 4 and 5A and S2 Table).
(A–C) Venn diagrams showing a five-fold or greater difference in expression at 1 (A), 2 (B) or 3 (C) days after NaCl treatment. Red numeric characters show the number of upregulated genes. Blue numeric characters show the number of downregulated genes.
(A) Venn diagrams showing a five-fold or greater difference in expression with all possible regressions under saline conditions (100 mM NaCl). Red numeric characters show the number of upregulated genes. Blue numeric characters show the number of downregulated genes. (B) Analysis of the gene ontology of 48 genes increased only in Krichauff under saline conditions. Functional categorizations by annotation were shown as gene ontology of biological process. (C) Analysis of the gene ontology of 21 genes increased only in Drysdale under saline conditions.
Krichauff consistently showed the highest STI compared with the other three varieties under moderate salinity stress conditions (Table 1). Transcriptome analysis showed that 87 genes were upregulated only in the leaf sheaths of Krichauff under salinity conditions (Fig 5A). The annotated nucleotide database of RAP-DB contained rice homologs of 48 of the 87 genes (Table 4). Using gene ontology analysis, most of the 48 genes were assigned as encoding proteins involved in cellular and metabolic processes (Fig 5B). The Krichauff-specific genes were accordingly classified as genes associated with the persistence of vigorous growth under mild salinity stress conditions.
Drysdale showed second highest STI compared with the other three varieties (Table 1). GO analysis showed that rice homologs of 21 of the 39 genes which were upregulated only in Drysdale were also classified as genes associated with the cellular, metabolic and biological processes (Fig 5C and Table 5).
To identify differences between gene expression profiles associated with shoot biomass development under control and saline conditions, gene sets showing expression only in Berkut under control conditions (Table 2 and S3 Table), only in Krichauff under saline conditions (Table 4 and S4 Table) and only in Drysdale under saline conditions (Table 5 and S5 Table) were compared. However, the Venn diagram shows no overlap among the 39 genes upregulated in Berkut, the 87 genes upregulated in Krichauff and the 39 genes upregulated in Drysdale (Fig 6A). The 58 genes downregulated in Berkut, the 124 genes downregulated in Krichauff and the 4 genes downregulated in Drysdale also showed no overlap (Fig 6B).
(A) Venn diagrams show the comparative analysis of upregulated gene expression associated with each specific physiological trait. (B) Venn diagrams show the comparative analysis of downregulated gene expression associated with each specific physiological trait.
Quantitative RT-PCR (qRT-PCR) was performed to validate the gene expression profiles obtained by microarray analysis. We randomly selected several genes which were expressed in either control or 100 mM NaCl treated plants (Figs 3A and 5A). The scatter plots showed that the results of the qRT-PCR analysis correlate highly with those of the microarray analysis (S5 Fig). The determination coefficient (R2 = 0.8418) indicates that the microarray analysis provides the reliable data to identify genes responding to the early stages of salinity stress.
Comparative analysis of gene expression profiles of bread wheat with those of Arabidopsis and rice revealed physiological differences in salinity stress responses
To determine whether the upregulated genes identified only in Krichauff and only in Drysdale under salinity treatment are also expressed in other plants, the expression profiles of the homologous genes in Arabidopsis and rice were examined using public gene expression databases (Tables 3–5). Using the Arabidopsis eFP Browser and Rice Expression Profile Database (RiceXPro), it was possible to determine the expression patterns of the 48 homologous of the upregulated Krichauff genes in Arabidopsis or rice after they had been exposed to either abiotic stress or hormone treatment. Only 12 rice and/or Arabidopsis homologs showed a tendency to respond to salinity stress and/or abscisic acid-inducible expression. In the 21 homologous genes upregulated only in Drysdale, the 7 rice and/or Arabidopsis homologs showed a tendency to response to stress-inducible expression. These results suggest that the gene expression responses in commercial bread wheat under mild salinity stress may be different from those in Arabidopsis and rice under severe salinity stress, and that bread wheat varieties with high RGR and high salinity tolerance have a unique transcriptome pattern.
Discussion
Several genes resolved in our microarray results have been reported as regulators of stress tolerance in other plants. Dehydrin/responsive to ABA (RAB), the homolog of CK213509 and CV761504, is a well-known family of hydrophilic proteins, and strongly responsive to water stresses and ABA treatment [28]. The cell wall-associated receptor kinase (WAK) and WAK-like kinase (WAKL) gene family are involved in biotic and abiotic stress responses and are required for cell elongation and development [29, 30]. The expression of WAKL4, the homolog of CJ851704, is induced by several mineral responses including to salt. This finding indicates that these membrane-to-cytoplasm communication signals are present in wheat under salinity stress as well as in other plants [31]. Lipid transfer protein (LTP) is involved in stress responses, and overexpression of LTP mediates salinity tolerance [32]. The LTP homolog AK332187 is expressed under salinity stress in stress-resistant wheat cultivars, suggesting a process of regulation similar to that for equilibrating metabolite accumulation. In addition, the results from the temporal microarray analysis performed in our study indicate that the experimental system properly captured the early response of gene expressions during the osmotic phase.
Our analysis sheds light on the unique molecular mechanisms of gene expression in commercial bread wheat under mild salinity conditions. The results from the 3D PCA revealed a wide diversity in leaf sheath transcriptome profiles for the four commercial cultivars studied (S4 Fig). Our results indicate that the salt stress-inducible gene sets generated during the osmotic phase are distinct between the four cultivars (Fig 5). The comparative analysis of variation in gene expression in relation to physiological function revealed the causal components associated with development of biomass and tolerance to mild salinity stress (Fig 6). In addition, most genes expressed in the leaf sheath of the four bread wheat cultivars tested do not correspond with salinity stress-inducible genes described in other well-characterized plants, such as rice and Arabidopsis (Tables 3–5). These results suggest that genetic variation resulting from selection during breeding exerts a strong influence on the development of genome architecture and adaptation to salinity conditions.
An integrated analysis of the transcriptomes and epigenomes in two rice subspecies revealed a high correlation of allelic bias of epigenetic modification with gene expression [33]. Bioinformatics approaches have shown that cis-regulatory mutations are the major source of evolutionary innovations and that these alterations lead to changes in gene regulatory information and phenotypic defects [34]. A recent genome-wide association study identified sequence variants between cultivars and wild rice in promoter regions [35]. Comparative transcriptome analysis between domesticated and wild tomatoes revealed that thousands of gene expression patterns were different and that most of them were associated with environmental responses and stress tolerance [36]. These studies are consistent with the present study in suggesting that genomic variation affects salinity tolerance through changes in gene expression.
In particular, gene ontology analysis revealed that the genes involved in a number of cellular and metabolic processes were more upregulated for shoot biomass development than stress response genes (Fig 5), indicating that salinity-tolerant cultivars have the ability of improving biomass development with the alterations in metabolism-mediated gene expressions compared with stress tolerance under mild salinity conditions. For Australian bread wheat, improvements in salinity tolerance and the ability to continue to increase biomass under mild salinity stress conditions are particularly important. The breeding of salinity tolerant cultivars can exploit this study for knowledge of an understanding of sequence variants associated with the domestication of salinity tolerant Australian bread wheat.
In this study, we evaluated, under control and mild saline conditions in a greenhouse, the physiological differences between four commercial bread wheat cultivars. The high-throughput phenotyping system at The Plant Accelerator of the Australian Plant Phenomics Facility allowed the measurement through time of shoot biomass, thus enabling calculations of RGR (Fig 1, S3 Fig and Table 1). This phenotyping analysis in The Plant Accelerator was more quantitative than is practically achievable in most other experiments in both controlled conditions and in the field. Therefore, the phenotyping system facilitates the analysis of useful germplasm using detailed molecular analyses, including microarrays. This study established that a high-throughput greenhouse phenotyping system, such as The Plant Accelerator, allows a powerful approach, combining physiological trait evaluation with molecular analysis of gene expression in bread wheat cultivars under mild saline conditions.
Our microarray analysis indicated that changes in the expression of salt stress-inducible genes reproducibly decreases within 2 days (Fig 4), suggesting that rapidly-induced gene expression, occurring shortly after the perception of stress, has a significant influence on adaptation to mild salinity conditions in the early osmotic phase. When combined with previous studies on Na+ and Cl- ions accumulation which gradually occurs under stress conditions, these results indicate that two phases, osmotic and ionic, are involved in responses and adaptation to mild salinity stress conditions in the bread wheat cultivars tested, and that the physiological characteristics of each phase are different from the perspective of acquiring salinity tolerance. Combined analysis of relative shoot growth, stress tolerance and Na+ exclusion constitutes a detailed analysis of mechanisms underlying salinity tolerance in both the osmotic and ionic phases during mild salinity treatment.
In conclusion, our integrated systematic analyses of phenomics and transcriptomics of various cultivars can provide useful information for a better understanding of the physiological and molecular processes involved in mild salinity tolerance in Australian bread wheat cultivars. The high-throughput phenotyping system at The Plant Accelerator of the Australian Plant Phenomics Facility has permitted a systematic analysis of the transcriptome profiles of bread wheat plants under well-controlled conditions and shed light on mild salinity-tolerance mechanisms represented by gene expression in leaf sheath during the early “osmotic” phase of salinity response and adaptation.
Supporting Information
S1 Table. Primer sets for quantitative RT-PCR analysis.
https://doi.org/10.1371/journal.pone.0133322.s001
(XLSX)
S2 Table. Differentially downregulated genes during salinity stress in Drysdale and Krichauff when compared to Berkut and Gladius.
Listed are genes whose expression was reduced more than five-fold in Drysdale and Krichauff compared with the other cultivars under salinity stress conditions. p values of <0.01 were included. Asterisk indicates stress- and/or abscisic acid treatment-inducible gene in rice and Arabidopsis. aProbe name as given by Agilent 44K wheat gene expression array (Design ID: 022297). bLocus indicates GenBank accession. cLoad ID as given by RAP-DB. dAGI, Arabidopsis Genome Initiative. eDescription as given by The Institute for Genomic Research database.
https://doi.org/10.1371/journal.pone.0133322.s002
(XLSX)
S3 Table. Differentially downregulated genes during biomass accumulation under control conditions in Berkut.
Listed are genes whose expression was reduced more than five-fold in Berkut compared with the other cultivars under control conditions. p values of <0.01 were included. Asterisk indicates stress- and/or abscisic acid treatment-inducible gene in rice and Arabidopsis. aProbe name as given by Agilent 44K wheat gene expression array (Design ID: 022297). bLocus indicates GenBank accession. cLoad ID as given by RAP-DB. dAGI, Arabidopsis Genome Initiative. eDescription as given by The Institute for Genomic Research database.
https://doi.org/10.1371/journal.pone.0133322.s003
(XLSX)
S4 Table. Differentially downregulated genes during salinity stress in Krichauff.
Listed are genes whose expression was reduced more than five-fold in Krichauff compared with the other cultivars under salinity stress conditions. p values of <0.01 were included. Asterisk indicates stress- and/or abscisic acid treatment-inducible gene in rice and Arabidopsis. aProbe name as given by Agilent 44K wheat gene expression array (Design ID: 022297). bLocus indicates GenBank accession. cLoad ID as given by RAP-DB. dAGI, Arabidopsis Genome Initiative. eDescription as given by The Institute for Genomic Research database.
https://doi.org/10.1371/journal.pone.0133322.s004
(XLSX)
S5 Table. Differentially downregulated genes during salinity stress in Drysdale.
Listed are genes whose expression was reduced more than five-fold in Drysdale compared with the other cultivars under salinity stress conditions. p values of <0.01 were included. Asterisk indicates stress- and/or abscisic acid treatment-inducible gene in rice and Arabidopsis. aProbe name as given by Agilent 44K wheat gene expression array (Design ID: 022297). bLocus indicates GenBank accession. cLoad ID as given by RAP-DB. dAGI, Arabidopsis Genome Initiative. eDescription as given by The Institute for Genomic Research database.
https://doi.org/10.1371/journal.pone.0133322.s005
(XLSX)
S1 Fig. Experimental layout with wheat varieties referred to the first experiment.
https://doi.org/10.1371/journal.pone.0133322.s006
(TIF)
S2 Fig. Experimental layout with wheat varieties referred to the second experiment.
https://doi.org/10.1371/journal.pone.0133322.s007
(TIF)
S3 Fig. Sequential monitoring of shoot biomass under conditions of salinity stress.
At the emergence of leaf 4, seedlings of (A) Berkut (blue diamonds, SE, n = 8), (B) Krichauff (orange triangles, SE, n = 8), (C) Gladius (green squares, SE, n = 8) or (D) Drysdale (red circles, SE, n = 7) cultivars grown with no added NaCl (open) or treated with 75 mM NaCl (filled), and digital images captured with an RGB camera. (E) Exponential curves were fitted to the data and relative growth rate (RGR) calculated for the four cultivars.
https://doi.org/10.1371/journal.pone.0133322.s008
(TIF)
S4 Fig. Clustering analysis of variation in gene expression under salinity stress for four cultivars.
(A) Hierarchical clustering of genes up- or downregulated with or without salinity stress in the four cultivars. Microarray analysis was performed at three time points (1, 2 or 3 days after treatment), and gene expression profiles with a five-fold or greater difference in expression were integrated for the treatment period. The condition tree of hierarchical classifications was performed with k-means of clustering algorithms. The Euclidean distances were calculated with the Manhattan method as the distance metric. (B) 3D PCA for the four cultivars. The gene expression clusters of Berkut (blue diamonds), Krichauff (orange triangles), Gladius (green squares) or Drysdale (red circles) under control conditions (open) and under salinity stress (filled) were shown. The Tracy-Widom (TM) test was used to perform the principal component analysis with the GeneSpring GX software.
https://doi.org/10.1371/journal.pone.0133322.s009
(TIF)
S5 Fig. Confirmation of the microarray data by qRT-PCR analysis.
The scatter plot shows the relative gene expression ratio of log2 (the signal intensity under saline conditions / the signal intensity under control conditions) between qRT-PCR analysis (X axis) and microarray analysis (Y axis). The six types of figures indicate each wheat gene among four cultivars. The blue, orange, green, and red colors indicate the data from Berkut, Krichauff, Gladius, and Drysdale, respectively. Correlation coefficient (R2) was 0.8418.
https://doi.org/10.1371/journal.pone.0133322.s010
(TIF)
Acknowledgments
We thank all the members of The Plant Accelerator for technical assistance and helpful suggestions. This research was supported by a JSPS special travel grant award for young scientists (MEXT; to F.T.), by the Programme for Promotion of Basic and Applied Researches for Innovations in Bio-oriented Industry (BRAIN; to K.S.), by the Ministry of Agriculture, Forestry and Fisheries (MAFF), and by the RIKEN Biomass Engineering Program Division (to F.T. and K.S.). The Plant Accelerator is funded by grants from the Australian and South Australian governments; and salinity tolerance research is supported by grants from the Australian Research Council, the Grains Research Development Corporation and King Abdullah University of Science and Technology (KAUST). The authors thank Christina Morris for helpful editorial comments on the manuscript.
Author Contributions
Conceived and designed the experiments: FT JT BB SJR MT. Performed the experiments: FT JT CT BB. Analyzed the data: FT JT BB SJR. Contributed reagents/materials/analysis tools: BB MS KS MT. Wrote the paper: FT JT BB SJR KS MT.
References
- 1. Munns R, Tester M. Mechanisms of salinity tolerance. Annual review of plant biology. 2008;59:651–81. Epub 2008/05/01. pmid:18444910.
- 2. Yamaguchi T, Blumwald E. Developing salt-tolerant crop plants: challenges and opportunities. Trends in plant science. 2005;10(12):615–20. Epub 2005/11/11. pmid:16280254.
- 3. Genc Y, McDonald GK, Tester M. Reassessment of tissue Na(+) concentration as a criterion for salinity tolerance in bread wheat. Plant, cell & environment. 2007;30(11):1486–98. Epub 2007/09/28. pmid:17897418.
- 4. Munns R, James RA. Screening methods for salinity tolerance: a case study with tetraploid wheat. Plant and Soil. 2003;253:201–18.
- 5. Roy SJ, Tucker EJ, Tester M. Genetic analysis of abiotic stress tolerance in crops. Current opinion in plant biology. 2011;14(3):232–9. Epub 2011/04/12. pmid:21478049.
- 6. Fleury D, Jefferies S, Kuchel H, Langridge P. Genetic and genomic tools to improve drought tolerance in wheat. Journal of experimental botany. 2010;61(12):3211–22. Epub 2010/06/08. pmid:20525798.
- 7. Roy SJ, Negrao S, Tester M. Salt resistant crop plants. Curr Opin Biotechnol. 2014;26:115–24. Epub 2014/04/01. pmid:24679267.
- 8. Yamaguchi-Shinozaki K, Shinozaki K. Transcriptional regulatory networks in cellular responses and tolerance to dehydration and cold stresses. Annual review of plant biology. 2006;57:781–803. Epub 2006/05/04. pmid:16669782.
- 9. Shinozaki K, Yamaguchi-Shinozaki K. Gene networks involved in drought stress response and tolerance. Journal of experimental botany. 2007;58(2):221–7. Epub 2006/11/01. pmid:17075077.
- 10. Wang C, Deng P, Chen L, Wang X, Ma H, Hu W, et al. A wheat WRKY transcription factor TaWRKY10 confers tolerance to multiple abiotic stresses in transgenic tobacco. PloS one. 2013;8(6):e65120. Epub 2013/06/14. pmid:23762295; PubMed Central PMCID: PMC3677898.
- 11. Zhang H, Mao X, Jing R, Chang X, Xie H. Characterization of a common wheat (Triticum aestivum L.) TaSnRK2.7 gene involved in abiotic stress responses. Journal of experimental botany. 2011;62(3):975–88. Epub 2010/10/30. pmid:21030389; PubMed Central PMCID: PMC3022395.
- 12. Zhang H, Mao X, Wang C, Jing R. Overexpression of a common wheat gene TaSnRK2.8 enhances tolerance to drought, salt and low temperature in Arabidopsis. PloS one. 2010;5(12):e16041. Epub 2011/01/07. pmid:21209856; PubMed Central PMCID: PMC3012728.
- 13. Zhang H, Mao X, Wu X, Wang C, Jing R. An abiotic stress response gene TaSnRK2.7-B in wheat accessions: genetic diversity analysis and gene mapping based on SNPs. Gene. 2011;478(1–2):28–34. Epub 2011/01/25. pmid:21256935.
- 14. Zhang H, Mao X, Zhang J, Chang X, Jing R. Single-nucleotide polymorphisms and association analysis of drought-resistance gene TaSnRK2.8 in common wheat. Plant physiology and biochemistry: PPB / Societe francaise de physiologie vegetale. 2013;70:174–81. Epub 2013/06/19. pmid:23774379.
- 15. Zhang H, Mao X, Zhang J, Chang X, Wang C, Jing R. Genetic diversity analysis of abiotic stress response gene TaSnRK2.7-A in common wheat. Genetica. 2011;139(6):743–53. Epub 2011/06/04. pmid:21637995.
- 16. Tian S, Mao X, Zhang H, Chen S, Zhai C, Yang S, et al. Cloning and characterization of TaSnRK2.3, a novel SnRK2 gene in common wheat. Journal of experimental botany. 2013;64(7):2063–80. Epub 2013/05/01. pmid:23630328; PubMed Central PMCID: PMC3638835.
- 17. Mao X, Zhang H, Tian S, Chang X, Jing R. TaSnRK2.4, an SNF1-type serine/threonine protein kinase of wheat (Triticum aestivum L.), confers enhanced multistress tolerance in Arabidopsis. Journal of experimental botany. 2010;61(3):683–96. Epub 2009/12/22. pmid:20022921; PubMed Central PMCID: PMC2814103.
- 18. Deng X, Hu W, Wei S, Zhou S, Zhang F, Han J, et al. TaCIPK29, a CBL-interacting protein kinase gene from wheat, confers salt stress tolerance in transgenic tobacco. PloS one. 2013;8(7):e69881. Epub 2013/08/08. pmid:23922838; PubMed Central PMCID: PMC3726728.
- 19. Berger B, de Regt B, Tester M. Trait dissection of salinity tolerance with plant phenomics. Methods Mol Biol. 2012;913:399–413. Epub 2012/08/17. pmid:22895775.
- 20. Brien CJ, Berger B, Rabie H, Tester M. Accounting for variation in designing greenhouse experiments with special reference to greenhouses containing plants on conveyor systems. Plant Methods. 2013;9(1):5. Epub 2013/02/09. pmid:23391282; PubMed Central PMCID: PMC3630016.
- 21. Golzarian MR, Frick RA, Rajendran K, Berger B, Roy S, Tester M, et al. Accurate inference of shoot biomass from high-throughput images of cereal plants. Plant Methods. 2011;7:2. Epub 2011/02/03. pmid:21284859; PubMed Central PMCID: PMC3042986.
- 22. Rajendran K, Tester M, Roy SJ. Quantifying the three main components of salinity tolerance in cereals. Plant, cell & environment. 2009;32(3):237–49. Epub 2008/12/05. pmid:19054352.
- 23. Connolly MA, Clausen PA, Lazar JG. Preparation of RNA from plant tissue using trizol. CSH Protoc. 2006;2006(1). Epub 2006/01/01. pmid:22485509.
- 24. Genc Y, Oldach K, Gogel B, Wallwork H, McDonald GK, Smith AB. Quantitative trait loci for agronomic and physiological traits for a bread wheat population grown in environments with a range of salinity levels. Molecular Breeding. 2013;32:39–59.
- 25. Genc Y, Taylor J, Rongala J, Oldach K. A major locus for chloride accumulation on chromosome 5A in bread wheat. PloS one. 2014;9(6):e98845. Epub 2014/06/04. pmid:24893005; PubMed Central PMCID: PMC4043771.
- 26. Furbank RT, Tester M. Phenomics—technologies to relieve the phenotyping bottleneck. Trends in plant science. 2011;16(12):635–44. Epub 2011/11/15. pmid:22074787.
- 27. Berger B, Parent B, Tester M. High-throughput shoot imaging to study drought responses. Journal of experimental botany. 2010;61(13):3519–28. Epub 2010/07/28. pmid:20660495.
- 28. Welin BV, Olson A, Nylander M, Palva ET. Characterization and differential expression of dhn/lea/rab-like genes during cold acclimation and drought stress in Arabidopsis thaliana. Plant molecular biology. 1994;26(1):131–44. Epub 1994/10/01. pmid:7948863.
- 29. Lally D, Ingmire P, Tong HY, He ZH. Antisense expression of a cell wall-associated protein kinase, WAK4, inhibits cell elongation and alters morphology. The Plant cell. 2001;13(6):1317–31. Epub 2001/06/13. pmid:11402163; PubMed Central PMCID: PMC135583.
- 30. Wagner TA, Kohorn BD. Wall-associated kinases are expressed throughout plant development and are required for cell expansion. The Plant cell. 2001;13(2):303–18. Epub 2001/02/28. pmid:11226187; PubMed Central PMCID: PMC102244.
- 31. Hou X, Tong H, Selby J, Dewitt J, Peng X, He ZH. Involvement of a cell wall-associated kinase, WAKL4, in Arabidopsis mineral responses. Plant physiology. 2005;139(4):1704–16. Epub 2005/11/16. pmid:16286448; PubMed Central PMCID: PMC1310553.
- 32. Pitzschke A, Datta S, Persak H. Salt stress in Arabidopsis: lipid transfer protein AZI1 and its control by mitogen-activated protein kinase MPK3. Molecular plant. 2014;7(4):722–38. Epub 2013/11/12. pmid:24214892; PubMed Central PMCID: PMC3973493.
- 33. He G, Zhu X, Elling AA, Chen L, Wang X, Guo L, et al. Global epigenetic and transcriptional trends among two rice subspecies and their reciprocal hybrids. The Plant cell. 2010;22(1):17–33. Epub 2010/01/21. pmid:20086188; PubMed Central PMCID: PMC2828707.
- 34. Meireles-Filho AC, Stark A. Comparative genomics of gene regulation-conservation and divergence of cis-regulatory information. Current opinion in genetics & development. 2009;19(6):565–70. Epub 2009/11/17. pmid:19913403.
- 35. Huang X, Kurata N, Wei X, Wang ZX, Wang A, Zhao Q, et al. A map of rice genome variation reveals the origin of cultivated rice. Nature. 2012;490(7421):497–501. Epub 2012/10/05. pmid:23034647.
- 36. Koenig D, Jimenez-Gomez JM, Kimura S, Fulop D, Chitwood DH, Headland LR, et al. Comparative transcriptomics reveals patterns of selection in domesticated and wild tomato. Proceedings of the National Academy of Sciences of the United States of America. 2013;110(28):E2655–62. Epub 2013/06/28. pmid:23803858; PubMed Central PMCID: PMC3710864.