Genome wide expression profiling of two accession of G. herbaceum L. in response to drought
- Alok Ranjan†1,
- Deepti Nigam†1,
- Mehar H Asif1,
- Ruchi Singh1,
- Sanjay Ranjan1,
- Shrikant Mantri2,
- Neha Pandey1,
- Ila Trivedi1,
- Krishan Mohan Rai1,
- Satya N Jena1,
- Bhupendra Koul1,
- Rakesh Tuli2,
- Uday V Pathre1 and
- Samir V Sawant1Email author
© Ranjan et al; licensee BioMed Central Ltd. 2012
Received: 12 September 2011
Accepted: 16 March 2012
Published: 16 March 2012
Genome-wide gene expression profiling and detailed physiological investigation were used for understanding the molecular mechanism and physiological response of Gossypium herbaceum, which governs the adaptability of plants in drought conditions. Recently, microarray-based gene expression analysis is commonly used to decipher genes and genetic networks controlling the traits of interest. However, the results of such an analysis are often plagued due to a limited number of genes (probe sets) on microarrays. On the other hand, pyrosequencing of a transcriptome has the potential to detect rare as well as a large number of transcripts in the samples quantitatively. We used Affymetrix microarray as well as Roche's GS-FLX transcriptome sequencing for a comparative analysis of cotton transcriptome in leaf tissues under drought conditions.
Fourteen accessions of Gossypium herbaceum were subjected to mannitol stress for preliminary screening; two accessions, namely Vagad and RAHS-14, were selected as being the most tolerant and most sensitive to osmotic stress, respectively. Affymetrix cotton arrays containing 24,045 probe sets and Roche's GS-FLX transcriptome sequencing of leaf tissue were used to analyze the gene expression profiling of Vagad and RAHS-14 under drought conditions. The analysis of physiological measurements and gene expression profiling showed that Vagad has the inherent ability to sense drought at a much earlier stage and to respond to it in a much more efficient manner than does RAHS-14. Gene Ontology (GO) studies showed that the phenyl propanoid pathway, pigment biosynthesis, polyketide biosynthesis, and other secondary metabolite pathways were enriched in Vagad under control and drought conditions as compared with RAHS-14. Similarly, GO analysis of transcriptome sequencing showed that the GO terms responses to various abiotic stresses were significantly higher in Vagad. Among the classes of transcription factors (TFs) uniquely expressed in both accessions, RAHS-14 showed the expression of ERF and WRKY families. The unique expression of ERFs in response to drought conditions reveals that RAHS-14 responds to drought by inducing senescence. This was further supported by transcriptome analysis which revealed that RAHS-14 responds to drought by inducing many transcripts related to senescence and cell death.
The comparative genome-wide gene expression profiling study of two accessions of G.herbaceum under drought stress deciphers the differential patterns of gene expression, including TFs and physiologically relevant processes. Our results indicate that drought tolerance observed in Vagad is not because of a single molecular reason but is rather due to several unique mechanisms which Vagad has developed as an adaptation strategy.
Drought is a major abiotic stress that affects plant growth and reduces plant yield. Many plants have evolved specific adaptive mechanisms in response to the drought stress exhibiting either drought escape or drought-tolerant mechanisms. Drought tolerance is a polygenic trait that involves a cascade of responses ranging from physiological changes to transcriptional regulation. The adaptive mechanisms in response to drought in plants are reduced water loss, reduced radiation absorption, reduced evaporation surfaces, and low tissue water potential. Other adaptive mechanisms include maintaining cell turgor pressure and reduced water loss by the accumulation of compatible solute molecules such as betaine, proline, sorbitol, and so on . The known molecular adaptive mechanisms in response to drought in plants involve the activation of transcriptional regulators such as DREB/CBF, MYB, and MYC [2, 3]. The development of microarray-based expression profiling methods have triggered significant progress in the characterization of the plant response to various abiotic stresses. Some recent efforts include identification of the gene networks involved in response to high-temperature stress in developing barley caryopses . Further, the cross-hybridization studies using Rice Gene chip lead to the identification of drought-inducible genes in banana . The dynamics of gene networks that are functional at the reproductive stage in response to drought stress have recently been studied in contrasting barley genotypes . Another technology for transcriptome sequencing that uses Roche's GS-FLX pyrosequencer leads to the identification of several unique genes in rice seedlings exposed to drought stress . Further, pyrosequencing helps in the identification of small RNA , microsatellite markers , and important agronomic traits [9–11] involved in the adaptation of plants to various abiotic stress conditions. Taken together, microarray and pyrosequencing have enormously contributed to the advancement of knowledge related to various genetic networks involved in adaptation.
Cotton (Gossypium sp.) is a leading textile fiber as well as the second most important oil seed crop in the world , and its productivity is vulnerable to drought. Molecular studies on cotton in response to drought are scanty. In drought-prone areas in Asia, diploid species Gossypium herbaceum (A1-genome) and Gossypium arboreum (A2-genome) are cultivated preferentially due to their inherent ability to withstand drought . In the NCBI database, till date, only 662 partial nucleotide sequences and 268 ESTs have been deposited for G. herbaceum, which is considerably low considering the genome size of G. herbaceum, which is 1.7 Gbp . The present study was undertaken to explore the G. herbaceum species with the aim of understanding the different genomic and physiological responses that might be involved in the inherent ability of this species to adapt to drought stress.
Screening of G. herbaceum accessions for drought tolerance and sensitivity
Fourteen accessions (Vagad, Gujcot21, RAHS-14, RAHS-IPS 187, H-17, AH-7GP, AH-127, RAHS 127, AH41, DB3-12, RAHS131, Jayelehar, GH18-2LC, and RAHS 132) of G. herbaceum that had been collected from different geographical locations in India were analyzed for drought tolerance and sensitivity based on mannitol-imposed drought stress in the DT1 experiment. Fifty seeds were maintained for germination, each at 2%, 4%, 6%, and 8% of mannitol in MS media of fourteen accessions. Accessions that grew and survived in 8% mannitol were considered drought tolerant, and other accessions that grew in only 4% mannitol were considered drought sensitive. In the DT2 experiment, two-week-old seedlings were exposed to drought by withholding water for seven days (soil moisture below 30%), whereas the control pots were irrigated daily. In the DT3 experiment, drought stress was given to the plants by withholding watering till soil moisture reaches below 30% in pots and drooping effects on plant leaves became prominent.
Vagad and RAHS-14 seedlings were grown during the summer of 2010 under field conditions at the Mahatma Gandhi Mission, Aurangabad, at Padegaon, India (19° 15' N; 75° 23' E, 513 m above sea level), in the rainout experimental station. Ten replicates of both the accessions were grown in three plots each with a surface area of 36 m2, filled with 1 m deep black cotton soil. A mobile greenhouse-grade polyethylene roof covered the plants during rainfall, and water was supplied through a trickle irrigation system in each plot. Plants were grown using otherwise normal agricultural practices, that is, plant spacing, fertilizers, and so on. Control plots were irrigated to field capacity twice weekly. Drought treatment was imposed by withholding irrigation. The measurements of different parameters were made after ten days (moderate stress) and after thirty days (severe stress) while withholding water. All gas exchange measurements were made between 8 h and 12 h, as gas exchange and environmental parameters did not show significant diurnal variation during this period.
Measurement of physiological parameters
Leaf gas exchange net photosynthetic rate (A), stomatal conductance (gs), and transpiration rate (E) were measured with an LI-6400 portable photosynthesis system (Li-Cor, Lincoln, NE) with red and blue LED light sources. All measurements were made on the third and fourth leaves from the terminal bud of a twig. Measurements were made between 8 h and 10 h on five leaves per treatment per accession on different plants. Measurements were conducted at c. 400 μmol CO2 mol_1 air, constant leaf temperature (T leaf = 33 ± 2°C), and constant vapor pressure deficit (VPD = 2.5 kPa ± 0.2) after the attainment of steady-state photosynthetic rates. The ratio of (A) to (E) was taken as the intrinsic photosynthetic water use efficiency (WUE). Dark respiration (R) was measured under similar microclimatic conditions after dark adaptation of the leaf for more than 30 min. Measurements of water potential and relative water content (RWC) were made at predawn on single, fully expanded leaves (third and fourth leaves from the terminal bud of a twig) immediately after excision. Leaf water potential (Ψ) was measured with a plant water status console (Soilmoisture, Santa Barbara, CA), whereas RWC of the leaf was calculated as 100 × (fresh weight - dry weight)/(turgid weight - dry weight) .
Sample collection and RNA isolation
Two accessions (Vagad and RAHS-14) of G.herbaceum were used for this study. Drought stress was given to potted plants by withholding water to maintain the soil moisture always less than 30%. The drought treated plants were watered only once in every alternate week while the control pots were irrigated daily. A total 12 plants were grown in earthen pots, including six plants from each accession. Drought stress, was given to plants by withhold watering in six pots including three from both accession. The drought treatment was given till the visible differences became apparent. Remaining six pots including three pots from both accessions were watered normally and considered as control. Thus, three plants from each accession at given condition were considered as biological replicates. Total RNA were extracted from the leaf tissues using Spectrum plant total RNA Kit, (Sigma-Aldrich) according to the manufacturer's instructions. After DNaseI treatment (Ambion), RNA were quantified and checked for the integrity by using a Bioanalyzer 2100 (Agilent, Inc., Palo Alto, CA, USA).
RNA labeling and hybridization
The direct labeling procedure was used with 1 μg of total RNA sample; double-stranded cDNA was synthesized with a T7 promoter-containing oligo (dT) primer using a Gene chip one-cycle cDNA synthesis kit (Affymetrix), followed by in vitro transcription using a Gene chip IVT labeling kit (Affymetrix). The biotinylated cRNA was fragmented for hybridization to Affymetrix cotton genome arrays and incubated at 45°C temperature for 16 h at 60 RPM in a hybridization oven. Arrays were washed and stained on an Affymetrix Fluidics Station 450. The arrays were scanned using Gene chip Scanner 3000. A summary of the image signal data for every gene interrogated on the array was generated using the Affymetrix MAS 5.0 (GCOS v1.3) statistical algorithm.
Microarray data analysis
We used Affymetrix Cotton Gene chip and Array Assist Software 5.2.2 (Agilent Technologies, Santa Clara, CA, USA) for comparative gene expression analysis. Raw cel files were exported from GCOS® software using data transfer tools for data processing and analysis in MeV and Array Assist Software 5.2.2 (Agilent Technologies, Santa Clara, CA, USA). Gene expression data analyses were completed using a filtered RMA expression value . Missing values were filtered, normalized, and natural log2 transformed for biological replicates. The t-test was used to determine the statistical significance of the differentially expressed gene. Probe IDs with detection p value ≤ 0.05 in three biological replicates were considered as present. Expression of genes in watered condition was compared between Vagad and RAHS-14 at p value ≤ 0.05 and fold Change (FC) ≥ 2.0. Similarly under drought stress condition expressed genes were analyzed between Vagad and RAHS-14. We have compared microarray data of Vagad and RAHS-14 in control and drought condition. Thus, when we indicate the genes as uniquely expressed in Vagad that means they were up-regulated in Vagad as compared to RAHS-14 and thus those genes were down regulated in RAHS-14 and vice a versa. The cRNA hybridization data were submitted according to MIAME guidelines, which were accessible through GEO series accession number GSE26522 http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE26522. The statistical analyses were conducted by MeV and Array Assist .
Annotation analyses of cotton Gene chip
Differentially up-regulated genes were analyzed using the functional categorization based on three GO categories at p-values ≥ 0.05. The agriGO tool http://bioinfo.cau.edu.cn/agriGO/ was used to perform the enrichment analysis using SEA (Singular Enrichment Analysis) coupled with available background data of cotton probes. Gene percentage analysis was calculated for each agriGO annotation in the GO category. Cotton Gene chip annotation was based on the top hits against the Arabidopsis genome (from TAIR release 8) using the PLEXdb tool and the Arabidopsis Genome Initiative databases.
Double-strand cDNA library preparation for GS-FLX pyrosequencing
Total RNA (3 μg) from apical leaf tissue from both the accessions were reverse transcribed using a T7-Oligo (dT) Promoter Primer in the first-strand cDNA synthesis (Affymetrix). After RNase H-mediated second-strand cDNA synthesis, the double-stranded cDNA was enriched and served as a template in the subsequent in vitro transcription (IVT) reaction (Affymetrix). The IVT reaction was carried out in the presence of T7 RNA Polymerase (Affymetrix). The cRNA (3 μg) was reverse transcribed in the first-strand cDNA synthesis step by using a random hexamer primer, followed by RNase H-mediated second-strand cDNA synthesis in replicates. The replicate samples were pooled and purified by the QIAquick PCR purification column (Qiagen) and the purified samples were used for sequencing.
Emulsion-based clonal amplification and pyrosequencing
Double-strand cDNA was nebulized in a fragment size between 400 and 600 bp. The fragmented cDNA were amplified in aqueous droplets that were made through the creation of a PCR reaction mixture in emulsion oil. The droplets act as separate microreactors in which parallel DNA amplifications are performed while yielding approximately 107 copies of a template per bead. One microliter of emulsion containing approximately 1.8 thousand beads was prepared. After PCR, the emulsion was broken to release the beads containing the amplified DNA template. The beads carrying the templates were enriched and deposited by centrifugation into open wells of a 70 × 70 mm2 optical picotiter plate. The beads containing a mixture of ATP sulfurylase and luciferase were loaded on the plates to generate light from free pyrophosphate to create the individual sequencing reactors in wells. The light generated from free pyrophosphate was transmitted through the base of the optical picotiter plate and detected by a large-format CCD. The images were processed to yield the sequence information.
Assembly and annotation of transcriptomes
All sequence analyses was conducted using publicly available software, such as R package http://www.R.project.org, MeV, and custom perl scripts. The quality-filtered reads were assembled at criteria of overlap size 100 bp and percent identity 96% using the CAP3 program. To remove the redundancy within both libraries, blastN was used in both libraries against itself, and the pooled sequences had ≥ 90% identity over the length of 75%. Only the largest sequence in each of these pools was considered. Using these criteria, the sequences obtained were called exemplars. The exemplar sequences for both libraries were tagged with the library name and pooled for annotation. For annotation, the pooled exemplars were queried against the NCBI nucleotide database (NT) using blastN at evalue of 10-10 and an alignment length of more than 50% of the query sequence. All the Gossypium ESTs available at the NCBI database were downloaded and pooled. The pooled exemplars were also queried against all public Cotton EST databases to identify new transcripts of Gossypium. Roche's GS-FLX sequence reads discussed in this article can be found in the Genebank http://www.ncbi.nlm.nih.gov/genbank of the National Center for Biotechnology with accession number SRA029162.
To assign the orientation of the transcripts, all the pooled exemplar sequences were analyzed by the ESTScan Model, which is trained on Arabidopsis and Oryza models. The sequences that passed the ESTScan model were translated according to the frame decided by the ESTScan program. These protein sequences were annotated using the blastP program against the NR Uniprot and pfam databases, at evalue of 10-10, and an alignment length of at least 50% of the query length. Gene names were assigned to each sequence based on the best blast hit.
The GO annotations for the sequences were derived using their Uniprot annotation. The Uniprot database was used, as it had extensive GO mapping. The GO annotation for level 5 was extracted for each library and used for further analysis.
Digital expression analyses
For the digital expression analysis, the reads for both libraries were tagged and pooled to form a large dataset of 141,722 reads. These reads were assembled using the CAP3 program at an overlap of 100 bp and 80% identity. These reads were assembled into 17,752 contigs. Further, the contigs were filtered to include only those that have more than five reads. We calculated the R statistics for the filtered genes to identify significant differentially expressing genes . To reduce the false discovery rate, only genes with an R value > 9 were considered. These filtered contigs were annotated using blastN against the NCBI nucleotide (NT) database, blastX against the NCBI non-redundant proteins (NR) and the Uniprot database.
The Quantitative Gene Expression (QGE) analyses
Recently, matrix-assisted lazer desorption ionization time of flight mass spectrometry (MALDI-TOF MS) was adopted for analyzing gene expression . Each PCR reaction was performed with 1 μl diluted cDNA (0.025 ng/μl), 0.5 μL 10x HotStar Taq PCR buffer, 0.2 μL MgCl2 (25 mM), 0.04 μL dNTP mix (25 mM each), 0.02 μL HotStar Taq Polymerase (50 U/μL, Qiagen), 0.1 μL competitor oligonucleotide (5 × 10-9 μM), 1 μL forward and reverse primer (1 μM each) (primers list, Additional file 1), and 2.14 μL ddH2O. The PCR condition was as follows: 95°C for 15 min for hot start, followed by denaturing at 94°C for 20 sec, annealing at 56°C for 30 sec, extension at 72°C for 1 min for 45 cycles, and finally, incubation at 72°C for 3 min. Excess dNTPs were removed from PCR products with shrimp alkaline phosphatase. A mixture of 0.17 μLhME buffer (SEQUENOM), 0.3 μL shrimp alkaline phosphatase (SEQUENOM), and 1.53 μL ddH2O was added to each PCR reaction. The reaction solutions were incubated at 37°C for 20 min, followed by 85°C for 5 min to inactivate the enzyme. Base extension reaction was performed by using 0.2 μL of selected ddNTPs/dNTP mixture (SEQUENOM), 0.108 μL of selected extension primer, 0.018 μL of ThermoSequenase (32 U/μL, SEQUENOM), and 1.674 μL ddH2O. The reaction mixture was kept at 94°C for 2 min, followed by 94°C for 5 sec, 52°C for 5 sec, and 72°C for 5 sec for 40 cycles. The extended reaction product was purified with spectroCLEAN resin (SEQUENOME) to remove salts in the buffer, and 16 μL resin/water solution was added into each base extension reaction. Approximately 10 nL of purified reaction product was dispensed onto a 384-format SpectroCHIP (SEQUENOM). A modified Bruker Biflex MALDI-TOF mass spectrometer was used for data acquisitions from the SpectroCHIP. Mass spectrometric data were automatically imported into the SpectroTYPER (SEQUENOM) database for automatic analysis such as noise normalization and peak area analysis.
Analysis of drought tolerance in G. herbaceum L. accessions
Screening of G. herbaceum accessions at different concentrations of mannitol.
Analyses of various physiological parameters in response to drought
Transcriptional profiling during drought and irrigated conditions in Vagad and RAHS-14
We used Affymetrix microarray and cotton chip for comparative expression profiling of leaves of irrigated and water-stressed plants of Vagad and RAHS-14 at a p value ≤ 0.05 and a fold change (FC) ≥ 2.0. We identified 656 and 535 genes as being differentially up-regulated in Vagad and RAHS-14, respectively, during the irrigated condition (Additional files 2 and 3). Similarly, 430 and 411 genes were differentially up-regulated in Vagad and RAHS-14, respectively, under the drought condition (Additional files 4 and 5). These differentially up-regulated genes were further annotated through the SEA (Singular Enrichment Analysis) method and identified enriched gene ontology (GO) terms. We considered only those GO terms that had been enriched at least twice over background cotton data and observed various significant differences in the metabolic pathways in both sensitive and tolerant accession.
Singular Enrichment Analysis (SEA) for identification of enriched GO terms in Vagad and RAHS-14 during irrigated condition
Singular Enrichment Analysis (SEA) for identification of enriched GO terms in Vagad and RAHS-14 during drought condition
Abiotic-stress-related process analysis of differentially expressed genes
RAHS-14 responds to drought by uniquely expressing ERFs (AP2/EREBP) and WRKY
Validation of identified significant genes by quantitative gene expression (QGE)
The transcriptome assembly and annotation
Summary of 454 transcriptome sequencing data generated for Vagad and RAHS-14 of G. herbaceum leaves transcriptome and assembly
Total reads (overlap size of 100 bp and 96% identity)a
Total contigs (100 bp or greater)b
Average length of contigs
Number of contigs with greater than 500 bp
cNumber of genes with significant hits in NCBI NR database
dNumber of genes with significant hits in cotton EST database
GO annotation of transcriptome
Differential gene expression analyses of transcriptomes
For differential expression analysis of the genes in both the transcriptome libraries, the reads from both the libraries were tagged and pooled to form one large dataset that was assembled into contigs using the CAP3 program (overlap length ≥ 100 bp and 80% identity). The 1,41,722 reads clustered into 17,752 contigs and resulted in 13,586 genes for the expression analysis. Significant changes in gene expression were calculated using R statistics (R value ≥ 9) (Additional file 14) and resulted in 2,026 genes, which seemed to be differentially regulated, and their differential expression was analyzed using Pearson uncentered correlation . Differentially expressed genes were annotated using the NCBI NR database (50% alignment length, evalue of 1 × 10-10). For each contig, the counts were converted to transcripts per million, which was transformed [log2 (fold change values)], and their ratio was calculated for fold changes between drought-tolerant and drought-sensitive tissues. A total of 275 contigs showed a very high expression in Vagad with nearly 49% (137) showing no hits to any proteins in the database (NR, NT, and Uniprot), and 10% genes were either hypothetical or putative expressed proteins, as they showed significant changes in expression. The remaining genes such as ascorbate peroxidase, cysteine protease, delta tonoplastic intrinsic proteins, LEA proteins, and so on that were related to drought stress were up-regulated (Additional file 15 worksheet 1). In RAHS-14, out of 484 genes, only 80 (15%) showed no hits to any protein or nucleotide in the database (NR and NT), and 18% were hypothetical proteins. The remaining 36% annotated genes were from the photosynthesis pathway, with a high expression of Rubisco activase, photosystem II D, and chlorophyll a\b binding proteins. The senescence-associated proteins constituted about 4% of the differentially expressed genes. Other up-regulated genes were cytochrome p450, heat shock protein 90, methionine synthase, and so on (Additional file 15 worksheet 2).
To analyze the differences in the drought tolerance of the G. herbaceum accessions at physiological and molecular levels, the plants were exposed to moderate and severe drought stress. It was observed that Vagad and RAHS-14 showed substantial differences in several physiological parameters and relative gene expression in response to drought. Vagad responded to moderate and severe drought by a gradual decline in A, gs, and E and, thus, has better WUE (Figure 2). In contrast, RAHS-14 had higher A and gs and, thus, lower WUE, which continued under moderate drought and then declined under severe stress (Figure 2a-d). In Vagad, earlier stomatal closure (Figure 2b) provides a large safety margin against embolism formation as observed in several species [20–22]; rather, plants are capable of pre-empting the water stress-induced xylem cavitation by closing their stomata [23–25]. Further, a reduction in A and a higher NPQ in Vagad accession under drought stress suggested that stress had been imposed and hypothesized that antioxidant defense systems and secondary metabolic pathways would have been enhanced in response. Indeed, our microarray expression analysis results showed that in Vagad, various primary and secondary metabolic pathways were enhanced (Figure 5). Flavonoid biosynthesis pathways especially lead to the formation of xanthophyll and finally convert it into anthocyanin, antheraxanthin, and zeaxanthin, which allows the quenching of excess energy from chlorophyll before it reaches reaction centers ; the transcript representing flavonoid biosynthesis pathways was found to be enriched in Vagad (Figure 5). In contrast, the xanthophyll-cycle conversion state was lower in RAHS-14 (Figure 2e and Additional file 7). Concurrent to physiological data, transcriptome analysis showed a higher expression of the vitamins metabolic process, such as α-tocopherol in Vagad (Additional file 13; worksheet 1). The α-tocopherol is a powerful antioxidant that scavenges and prevents the formation of free radicals and prevents lipid peroxidation, thereby resulting in damage to thylakoid and chloroplast membranes . Besides flavonoid pathways in Vagad, various other metabolic pathways, including polyketide biosynthesis, phenyl propanoid biosynthesis, and shikimate pathways, synthesize protective molecules such as stilbene, coumarins, and ligninis and are enriched in Vagad, which might help in the detoxification of free radicals and give an advantage to Vagad for surviving in drought stress (Figure 4). These secondary metabolites, particularly phenylpropanoid, were widely reported for their multiple function in response to various forms of environmental stress . The genes related to the shikimate and phenylpropanoid pathways have been reported to express a higher level in drought-tolerant tomato cultivars (Solanum lycopersicum L.) as compared with drought-sensitive tomato cultivars . In RAHS-14, the lipid metabolism processes were significantly higher (Figure 5). In water-deficit conditions, the membranes are the main targets of the degradative process, resulting in the formation of polar and non-polar lipid molecules . Thus, the higher lipid metabolic processes in RAHS-14 probably reflect higher membrane degeneration and, hence, the necessity of higher lipid metabolism for membrane homeostasis correlated with their susceptibility of drought stress. Vagad, in contrast to RAHS-14, maintained lipid metabolism and membrane integrity to resist the drought stress. RAHS-14 showed a higher expression of genes related to nucleic acid (DNA/RNA) metabolism, whereas the expression of these genes remains unchanged in Vagad (Additional file 5 and Figure 6). Higher nucleic acid metabolism in RAHS-14 might show an interesting mechanism related to the energy state of the cell that is represented by the salvage pathways . Phosphoribosyl-1-pyrophosphate, a key intermediate component for the synthesis of ribose-5-phosphate in nucleic acid metabolism, showed higher expression in RAHS-14 as representing the operation mechanism of salvage pathways and helps RAHS-14 keep the energy pools from being used up too quickly (Additional file 13; worksheet 2). It indicated that RAHS-14 responds to drought by inducing energy-consuming processes, whereas Vagad has various inherent primary and secondary metabolic processes that maintain growth, albeit slow even under drought. This is further evident from dark respiration (R) data which show that R declined in Vagad and increased in RAHS-14 during drought (Figure 2h). Many genes that were up-regulated in response to drought stress in Vagad are reported to be involved in multiple mechanisms that may contribute to drought tolerance (Additional files 4 and 15, worksheet 1). For example, Aquaporins (AQP), a water-selective channel protein, known to mediate and regulate rapid transmembrane water flow during a wide range of stress response, stomatal movement, and water channel movements, was expressed at a higher level in Vagad (Additional file 4) . Similarly, many tonoplastic intrinsic proteins (TIPs) from maize, Arabidopsis, and radish are known to control water exchange between cytosol and vacuole in salt stress and drought stress  and these genes were expressed at a higher level in Vagad (Additional file 4). DnaJ heat shock protein/chaperone and Delta 1-pyrroline-5-carboxylate synthase 2 (P5CS2), which enhance root biomass, flowering, and seed setting during abiotic stress , were expressed at a higher level in Vagad (Additional files 2 and 3). Yet another significantly induced gene in Vagad was the ERF/AP2-type transcription factor (RAP2.1), which was reported to be induced in drought and cold via an ABA-independent pathway . RAHS-14 expressed more senescence-related genes, whereas we found that in Vagad, genes related to alcohol dehydrogenase and late embryogenesis protein 5 was expressed at a higher level (Additional files 5 and 15). Previous studies indicate that the accumulation of late embryogenesis abundant proteins and alcohol dehydrogenase (Adh) gene are correlated with stress tolerance [35, 36]. The significant up-regulation of LEA and Adh genes in Vagad in drought stress suggests that these genes play an important role in conferring drought tolerance, whereas RAHS-14 tends toward the senescence during drought stress. Several transcription families were significantly and differentially expressed in both the accessions. Ethylene-responsive element binding factors (ERFs) are members of a novel family of transcription factors that are specific to plants and which regulate nuclear gene expression under various stress conditions. Six different members homologous to the Arabidopsis ERF family (At1g19210; At1g28360; At4g34410; At5g44210; At5g47220; At5g47230) were expressed exclusively in RAHS-14 under drought stress. The expression of ERF in RAHS-14 in the drought condition indicates that RAHS-14 responds to drought mainly by the ethylene pathway, thereby leading to senescence. This relates well with transcriptome data showing a higher level of senescence-related transcripts in RAHS-14. In Vagad, these AP2/EREBP TFs belong to the Cytokinin Response Factors (CRF2) class (At1g78080; At4g23750). CRFs function redundantly to regulate the various metabolic functions, including transpiration, stomatal conductance, and respiration [37, 38]. Further, in RAHS-14, seven (At1g29860; At1g80840; At2g23320; At2g24570; At2g38470; At3g56400; At4g24240) and four (At1g80840; At2g47260; At4g24240; At5g49520) WRKY transcription factors were found to be uniquely expressed in drought and irrigated conditions, respectively. Previous reports showed that abscisic acid and salicylic acid have been involved in the WRKY-mediated hormone signal pathway during abiotic and biotic stress . Their precise role in the abiotic stress response regulatory network is not fully understood . In contrast, the MYC-type bHLH transcription factor, which regulates the expression of CBF3/DREB1A in abiotic stress, was enriched in Vagad, which suggests the involvement of the MYC-type bHLH transcription factors in the expression of CBF/DREB1 genes in Vagad but less involvement in RAHS-14 . Interactions between CBF/DREB1 genes and bHLH TFs and their involvement in various oxidative-mediated processes justify their presumed roles as regulators of drought response in Vagad. Other members of TFs, such as bZIP (At2g46270; At4g34590; At5g24800) and GRAS families (At5g48150; At5g52510; At5g66770), were expressed in higher numbers in Vagad; the regulatory roles of these TFS have been reported in stress responses in plants. The GO-based analysis of both the transcriptome libraries revealed many metabolic processes and responses to various forms of abiotic stress that were specific to Vagad and RAHS-14 (Figure 11). The shift in the processes toward reproductive growth and senescence in RAHS-14 clearly showed that they had crossed the threshold of stress and were proceeding toward senescence. The up-regulation of several metallothioneins, lipid transfer proteins, lea proteins, sucrose synthase, and so on in Vagad showed the induction of defense and stress-related genes to combat drought stress. In addition, the down-regulation of many photosynthesis-related genes during drought stress (Additional file 15 worksheet 2) can be attributed to the conservation of energy for the survival in drought stress. The up-regulation of genes such as aquaporins, lea, and metallothioneins may have played a major role in imparting tolerance to Vagad.
We further compared differentially expressed genes obtained by microarray and contigs obtained in transcriptomic data using Pearson uncentered correlation. A total 167 differentially expressed genes obtained in microarray showed very high Pearson correlation coefficient of 0.845 with transcriptomic contigs obtained under drought stress condition (Additional file 16). Out of these 167 differentially expressed genes, 78 and 48 genes were uniquely represented in Vagad and RAHS-14 respectively under drought stressed condition. Uniquely represented genes under drought stress in Vagad were mainly involved in synthesis of membrane and cytoskeleton associated proteins, oxidoreductases, kinases, heat shock proteins, sugar alcohols and secondary metabolites like, cumurin and stilbene. Similarly in RAHS-14, these genes were mainly associated with transcription factors (viz, WRKY, AP2, WD40, Zinc finger and ERF), senescence associated proteins, ethylene and auxin responsive elements (Additional file 17). The GO based analysis of these 167 genes revealed many processes that were specific to Vagad and RAHS-14. The upregulation of pyrroline-5- carboxylase, ATPase, inositol etc. (Additional file 17) in Vagad may have played a major role in imparting drought tolerance. In RAHS-14 higher numbers of these genes were involved in senescence and ethylene mediated signaling which clearly indicate the RAHS- 14 crossed the threshold of drought stress and are proceeding towards senescence (Additional file 15). Thus both the method in an unbiased way identifies same mechanistic model operating for drought responsiveness in Vagad and RAHS 14.
We conclude that drought-tolerant accessions such as Vagad must have developed multiple mechanisms as adaptive behavior against drought. These mechanisms are interlinked and probably cannot be seen in isolation; understanding these mechanisms will be helpful for developing our future drought-tolerant varieties.
Complementary deoxyribonucleic acid
Complementary ribonucleic acid
False detection rate
Singular enrichment analysis
Expressed sequence tag
Quantitative gene expression
Charge couple device
Water use efficiency
Relative water content.
We thank Dr. S. S. Patil for providing accessions of G. herbaceum. AR and DN thanks to CSIR India for supporting fellowship as SRF. This work was supported under CSIR supra institutional projects SIP03 and SIP05.
- Seki M, Umezawa T, Urano K, Shinozaki K: Regulatory metabolic networks in drought stress responses. Curr Opin Plant Biol. 2007, 10 (3): 296-302. 10.1016/j.pbi.2007.04.014.View ArticlePubMedGoogle Scholar
- Bartels D, Sunkar R: Drought and salt tolerance in plants. Crit Rev Plant Sci. 2005, 24 (1): 23-58. 10.1080/07352680590910410.View ArticleGoogle Scholar
- Ingram J, Bartels D: The Molecular Basis of Dehydration Tolerance in Plants. Annu Rev Plant Physiol Plant Mol Biol. 1996, 47: 377-403. 10.1146/annurev.arplant.47.1.377.View ArticlePubMedGoogle Scholar
- Mangelsen E, Kilian J, Harter K, Jansson C, Wanke D, Sundberg E: Transcriptome analysis of high-temperature stress in developing barley caryopses: early stress responses and effects on storage compound biosynthesis. Molecular Plant. 2011, 4 (1): 97-115. 10.1093/mp/ssq058.View ArticlePubMedGoogle Scholar
- Davey MW, Graham NS, Vanholme B, Swennen R, May ST, Keulemans J: Heterologous oligonucleotide microarrays for transcriptomics in a non-model species; a proof-of-concept study of drought stress in Musa. BMC Genomics. 2009, 10 (1): 436-10.1186/1471-2164-10-436.PubMed CentralView ArticlePubMedGoogle Scholar
- Guo P, Baum M, Grando S, Ceccarelli S, Bai G, Li R, von Korff M, Varshney RK, Graner A, Valkoun J: Differentially expressed genes between drought-tolerant and drought-sensitive barley genotypes in response to drought stress during the reproductive stage. J Exp Bot. 2009, 60 (12): 3531-3544. 10.1093/jxb/erp194.PubMed CentralView ArticlePubMedGoogle Scholar
- Li YF, Zheng Y, Addo-Quaye C, Zhang L, Saini A, Jagadeeswaran G, Axtell MJ, Zhang W, Sunkar R: Transcriptome-wide identification of microRNA targets in rice. The Plant Journal. 2010, 62 (5): 742-759. 10.1111/j.1365-313X.2010.04187.x.View ArticlePubMedGoogle Scholar
- Wheat CW: Rapidly developing functional genomics in ecological model systems via 454 transcriptome sequencing. Genetica. 2010, 138 (4): 433-451. 10.1007/s10709-008-9326-y.View ArticlePubMedGoogle Scholar
- Wilkins TA, Arpat AB: The cotton fiber transcriptome. Physiol Plant. 2005, 124 (3): 295-300. 10.1111/j.1399-3054.2005.00514.x.View ArticleGoogle Scholar
- Barbazuk WB, Emrich SJ, Chen HD, Li L, Schnable PS: SNP discovery via 454 transcriptome sequencing. Plant J. 2007, 51 (5): 910-918. 10.1111/j.1365-313X.2007.03193.x.PubMed CentralView ArticlePubMedGoogle Scholar
- Varshney RK, Nayak SN, May GD, Jackson SA: Next-generation sequencing technologies and their implications for crop genetics and breeding. Trends Biotechnol. 2009, 27 (9): 522-530. 10.1016/j.tibtech.2009.05.006.View ArticlePubMedGoogle Scholar
- Wendel JF, Cronn RC: Polyploidy and the evolutionary history of cotton. Advances in Agronomy. 2003, 78: 139-186.View ArticleGoogle Scholar
- Kulkarni VN, Khadi BM, Maralappanavar MS, Deshapande LA, Narayanan SS: The Worldwide Gene Pools of Gossypium arboreum L. and G. herbaceum L., and Their Improvement. 2009, New York: springer, 3:Google Scholar
- Chen ZJ, Scheffler BE, Dennis E, Triplett BA, Zhang T, Guo W, Chen X, Stelly DM, Rabinowicz PD, Town CD: Toward sequencing cotton (Gossypium) genomes. Plant Physiol. 2007, 145 (4): 1303-1310. 10.1104/pp.107.107672.PubMed CentralView ArticlePubMedGoogle Scholar
- John S, Boyer RA: Passioura Osmotic adjustment leads to anomalously low estimates of relative water content in wheat and barley. Funct Plant Biol. 2008, 35: 1172-1182. 10.1071/FP08157.View ArticleGoogle Scholar
- Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. 2003, 4 (2): 249-264. 10.1093/biostatistics/4.2.249.View ArticlePubMedGoogle Scholar
- Eisen MB, Spellman PT, Brown PO, Botstein D: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA. 1998, 95 (25): 14863-14868. 10.1073/pnas.95.25.14863.PubMed CentralView ArticlePubMedGoogle Scholar
- Stekel DJ, Git Y, Falciani F: The comparison of gene expression from multiple cDNA libraries. Genome Res. 2000, 10 (12): 2055-2061. 10.1101/gr.GR-1325RR.PubMed CentralView ArticlePubMedGoogle Scholar
- Ding C, Cantor CR: A high-throughput gene expression analysis technique using competitive PCR and matrix-assisted laser desorption ionization time-of-flight MS. Proc Natl Acad Sci USA. 2003, 100 (6): 3059-3064. 10.1073/pnas.0630494100.PubMed CentralView ArticlePubMedGoogle Scholar
- Froux FHR, Ducrey M, Dreyer E: Xylem hydraulic efficiency versus vulnerability in seedlings off our contrasting Mediterranean tree species (Cedrusatlantica, Cupressus sempervirens, Pinushalepensis and Pinusnigra). Ann For Sci. 2002, 59: 409-418.View ArticleGoogle Scholar
- Martínez-Vilalta J, Prat E, Oliveras I, Piñol J: Xylem hydraulic properties of roots and stems of nine Mediterranean woody species. Oecologia. 2002, 133 (1): 19-29. 10.1007/s00442-002-1009-2.View ArticlePubMedGoogle Scholar
- Pockman WT, Sperry JS: Vulnerability to xylem cavitation and the distribution of Sonoran Desert vegetation. Am J Bot. 2000, 87 (9): 1287-1299. 10.2307/2656722.View ArticlePubMedGoogle Scholar
- Nardini A, Salleo S: Limitation of stomatal conductance by hydraulic traits: sensing or preventing xylem cavitation?. Trees-Struct Funct. 2000, 15 (1): 14-24. 10.1007/s004680000071.View ArticleGoogle Scholar
- Cochard H, Coll L, Le Roux X, Améglio T: Unraveling the effects of plant hydraulics on stomatal closure during water stress in walnut. Plant Physiol. 2002, 128 (1): 282-290. 10.1104/pp.010400.PubMed CentralView ArticlePubMedGoogle Scholar
- Brodribb TJ, Holbrook NM: Stomatal closure during leaf dehydration, correlation with other leaf physiological traits. Plant Physiol. 2003, 132 (4): 2166-2178. 10.1104/pp.103.023879.PubMed CentralView ArticlePubMedGoogle Scholar
- Park H, Kreunen SS, Cuttriss AJ, DellaPenna D, Pogson BJ: Identification of the carotenoid isomerase provides insight into carotenoid biosynthesis, prolamellar body formation, and photomorphogenesis. Plant Cell Online. 2002, 14 (2): 321-332. 10.1105/tpc.010302.View ArticleGoogle Scholar
- Gigon A, Matos AR, Laffray D, Zuily-Fodil Y, Pham-Thi AT: Effect of drought stress on lipid metabolism in the leaves of Arabidopsis thaliana (ecotype Columbia). Ann Bot. 2004, 94 (3): 345-351. 10.1093/aob/mch150.PubMed CentralView ArticlePubMedGoogle Scholar
- Dixon RA, Paiva NL: Stress-induced phenylpropanoid metabolism. The Plant Cell. 1995, 7 (7): 1085-1097.PubMed CentralView ArticlePubMedGoogle Scholar
- Sánchez-Rodríguez E, Moreno DA, Ferreres F, Rubio-Wilhelmi MM, Ruiz JM: Differential responses of five cherry tomato varieties to water stress: changes on phenolic metabolites and related enzymes. Phytochemistry. 2011, 72 (8): 723-729. 10.1016/j.phytochem.2011.02.011.View ArticlePubMedGoogle Scholar
- Dobrota C: Energy dependant plant stress acclimation. Rev Environ Sci Biotechnol. 2006, 5: 243-251. 10.1007/s11157-006-0012-1.View ArticleGoogle Scholar
- Gao Z, He X, Zhao B, Zhou C, Liang Y, Ge R, Shen Y, Huang Z: Overexpressing a Putative Aquaporin Gene from Wheat, TaNIP, Enhances Salt Tolerance in Transgenic Arabidopsis. Plant Cell Physiology. 2010, 51 (5): 767-775. 10.1093/pcp/pcq036.View ArticlePubMedGoogle Scholar
- Forrest KL, Bhave M: Major intrinsic proteins (MIPs) in plants: a complex gene family with major impacts on plant phenotype. Funct Integr Genomics. 2007, 7 (4): 263-289. 10.1007/s10142-007-0049-4.View ArticlePubMedGoogle Scholar
- Kishor P, Hong Z, Miao GH, Hu C, Verma D: Overexpression of [delta]-Pyrroline-5-carboxylate synthetase increases proline production and confers osmotolerance in transgenic plants. Plant Physiol. 1995, 108 (4): 1387-1394.PubMed CentralPubMedGoogle Scholar
- Zhu Q, Zhang J, Gao X, Tong J, Xiao L, Li W, Zhang H: The Arabidopsis AP2/ERF transcription factor RAP2.6 participates in ABA, salt and osmotic stress responses. Gene. 2010, 457 (1-2): 1-12. 10.1016/j.gene.2010.02.011.View ArticlePubMedGoogle Scholar
- Dalal M, Tayal D, Chinnusamy V, Bansal KC: Abiotic stress and ABA-inducible Group 4 LEA from Brassica napus plays a key role in salt and drought tolerance. J Biotechnol. 2009, 139 (2): 137-145. 10.1016/j.jbiotec.2008.09.014.View ArticlePubMedGoogle Scholar
- Li Y, Zhu Y, Liu Y, Shu Y, Meng F, Lu Y, Bai X, Liu B, Guo D: Genome-wide identification of osmotic stress response gene in Arabidopsis thaliana. Genomics. 2008, 92 (6): 488-493. 10.1016/j.ygeno.2008.08.011.View ArticlePubMedGoogle Scholar
- Dodd IC: Hormonal interactions and stomatal responses. J Plant Growth Regul. 2003, 22 (1): 32-46. 10.1007/s00344-003-0023-x.View ArticleGoogle Scholar
- Rivero RM, Shulaev V, Blumwald E: Cytokinin-dependent photorespiration and the protection of photosynthesis during water deficit. Plant Physiol. 2009, 150 (3): 1530-1540. 10.1104/pp.109.139378.PubMed CentralView ArticlePubMedGoogle Scholar
- Cheong YH, Chang HS, Gupta R, Wang X, Zhu T, Luan S: Transcriptional profiling reveals novel interactions between wounding, pathogen, abiotic stress, and hormonal responses in Arabidopsis. Plant Physiol. 2002, 129 (2): 661-677. 10.1104/pp.002857.PubMed CentralView ArticlePubMedGoogle Scholar
- Yun KY, Park M, Mohanty B, Herath V, Xu F, Mauleon R, Wijaya E, Bajic V, Bruskiewich R, de los Reyes B: Transcriptional regulatory network triggered by oxidative signals configures the early response mechanisms of japonica rice to chilling stress. BMC plant biology. 2010, 10 (1): 16-10.1186/1471-2229-10-16.PubMed CentralView ArticlePubMedGoogle Scholar
- Guo Y, Xiong L, Ishitani M, Zhu JK: An Arabidopsis mutation in translation elongation factor 2 causes superinduction of CBF/DREB1 transcription factor genes but blocks the induction of their downstream targets under low temperatures. Proc Natl Acad Sci. 2002, 99 (11): 7786-7791. 10.1073/pnas.112040099.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.