Drought Acclimation Studies in Peanut
Peanut is the second most important legume crop in the world just behind soybean. According to Food and Agriculture Organization (FAO), the annual global production of peanut is 51.7 million tons produced from 25.8 million hectares of cultivated land. Majority of the world peanut production (about 97%) is produced from Asia, Africa, and South America. China leads in production with a 37 % share in the global production, followed by India, USA, and Nigeria. The U.S. is one of the world’s leading peanut exporters, with average annual exports of between 200,000 and 250,000 metric tons with Argentina and China being other major exporters. Over 70% of peanut cultivation area is in arid and semi-arid climates where the plants are more prone to drought stress conditions (Kambriranda et. al -2011). So, drought studies in peanut are justified and are very important to maintain sustainable global production of peanut.
What is a drought?
Drought can be defined as the absence of rainfall or irrigation for a period of time sufficient to deplete soil moisture and cause dehydration in plant tissues, enough to interfere with its normal physiological processes. Drought stress has many adverse effects on plant growth and metabolism. Major drought effects include oxidative stress to disrupt photosynthetic apparatus, loss of cell wall turgor, disruption of Ion transport, Inhibition of cell growth and expansion, inhibition of stem growth etc.
Subjecting peanut plants to early season water deficit imparts late season tolerance to drought conditions. This effect is seen as the early water stress, primes the physiological and genetic responses of the plant to tolerate drought stress. So, this strategy of subjecting plants to water stress in the early season and fully irrigating in late season is called as primed acclimation (Rowland et. al 2012).
Water stress and high temperatures are the major factors affecting peanut production in the U.S. this condition can be attributed to either regional aridity or uneven rainfall pattern across the country (Rowland et. al 2012). There are a limited number of studies focused on peanut drought and acclimation responses due to the complexity of its genome and bioinformatic constraints like unavailability of a complete whole genome.
Drought response factors are widely studied in crop plants. Different drought response mechanisms have been identified in leaves and roots of crop plants which range from cell wall modifications to differential regulation of transcription factors. Cell wall being the first line of defense in plants show immediate response to drought stress by expression of genes which increase the thickness of the cell wall and maintaining its plasticity by increased production of expansin. In major crops plants, an upregulation of XTH and expansin genes have been reported under Water stress conditions which imply changes in cell wall architecture as a response to the stress induced (Zhu et al 2007).
An increase in cellulose synthesis has been reported by Zheng et al in 2014 which indicates that the plant is responding to drought stress by maintaining the cell turgor pressure. Deposition of pectin in the cell wall also aids in maintaining the cell wall integrity and increase its thickness to combat abiotic stress. Tubulin and Actin are also among the important proteins which are expressed during drought stress which regulate cytoskeleton re-arrangement.
Drought acclimation responses are different between roots and leaves. Leaves are the first areal organs which are exposed to heat and are also the seat of evapotranspiration due to the presence of stomata and having a large surface area. In leaves, drought stress generally induces stomatal closure and oxidative stress. It has been reported that oxidative stress negatively affects photosynthetic apparatus and thus most of the plants show reduced photosynthetic rate during the drought stress. Genes which code for heat shock proteins, transcription factors and transferases have shown to be involved in drought tolerance mechanisms of both roots and leaves.
Carbohydrate metabolism is another major aspect in drought acclimation response. Kottapalli et al has reported a reduction in sucrose mobility and starch synthesis and a prominent increase in starch degradation. This shows the cell’s response to drought stress by accumulating osmolytes within which act as osmoprotectants against drought stress. This also explains the smaller size of seeds and decreased biomass of the plants during drought stress. In leaves due to this phenomenon, we see a decreased content of complex carbohydrates like starch as they are broken down to simple carbohydrates like fructose and glucose as reported by Keller et. al.
Changes in root morphology has been attributed to adaptive responses towards drought stress. Based on the plant species, roots may show increased elongation or expression of genes which involve in the inhibition of cell elongation in the roots. Upregulation of genes coding for myosin, tubulin and callose synthase which involve in microtubule synthesis enhance the elongation of root cells. Response to drought stress in roots showed upregulation of signaling and transcription factors coupled with upregulation of abscisic acid and jasmonates.
Root Box Experiment
Hypothesis: During drought acclimation, there is upregulation of auxin signaling pathway that results in the induction of lateral root formation, whereas during drought stress, the lateral root formation is inhibited.
Aim: To validate the hypothesis, root box experiment was conducted to study the root morphology during drought stress and acclimation.
A root box experiment was designed to study the root morphology of peanut plants under different water stress conditions. C76-16, a drought tolerant national check variety was used in this experiment. A root box is a square box made of wood with a glass panel on one side to study the root morphology. The wooden side of the root box has several rows of nails to separate the roots. Root boxes were filled approximately 80% with sand and 20% with the potting mixture. These boxes are irrigated to saturation and seeds were sown after two to three hours. A total of six root boxes were taken, two for control, two for acclimation treatment and two for non-acclimation treatment. All the six boxes were fully irrigated during the first 15 days of the growth period. Control boxes were fully irrigated for the whole duration of the experiment while acclimated and non-acclimated boxes were subjected to different spells of drought stress. After the initial growth period, acclimated boxes were subjected to 50 percent water deficit for 14 days called as dry down 1 (DD1). Bot control and non-acclimated plants were fully irrigated during this period. After DD1, all the plants are fully irrigated for the next 14 days which is called the recovery period. Then in second dry down spell (DD2), both acclimated and non-acclimated plants were subjected to 50 percent water deficit. According to our hypothesis, acclimated plants should have better lateral root formation than the non-acclimated plants.
A Meta-Analysis of Drought Acclimation Responses Across Three Peanut Genotypes
Hypothesis: Different peanut genotypes have different responses to drought stress conditions.
Aim: To perform a meta-analysis of RNA-seq data from the leaf and root samples of three different peanut genotypes.
Three different peanut genotypes were used in this meta-analysis.
Drought acclimation experiments were conducted on all the three genotypes. Root and leaf samples were collected at various time points and mRNA was extracted from them. These samples were sequenced on Illumina Hiseq2500 sequencer in rapid mode and data sets were generated for each genotype. For the analysis, 7day stress timepoint was selected to compare across the genotypes. The first approach was creating a meta-transcriptome merging the reads from all the genotypes and then mapping the reads to that transcriptome to identify differentially expressed genes. The reads from all the three genotypes were concatenated on the local server and then those merged reads were used to create the reference transcriptome using a software called Trinity. Separate transcriptomes were made in the way explained above for root and leaf data across the three genotypes.
Creating a de novo Reference transcriptome :
The raw data from Illumina sequencing was in BCL format which was converted to zipped fastq format using BCL to fastq conversion pipeline on the local server. All the fastq files were unzipped on the local server using gunzip command. As the sequences from the different samples were concatenated to single R1 and R2 files for the de novo assembly. A reference transcriptome was created using Trinity software developed by Broad institute to assemble high quality reference transcriptomes on the local server. Trinity performs Insilco normalization of the reads and the assembly takes place in three steps namely inch worm, chrysalis and butterfly using the principle of debruijn graphs. An in -house python script was used to assign contig numbers starting from 1 all the way to the end of the trinity reference created.
Alignment of the reads to the reference and differential gene expression analysis.
All the unzipped fastq files and the trinity reference were loaded into Qseq, a module of DNAstar Navigator software to align the RNA-seq reads onto the trinity assembly. After the assembly is done, the reads that were uniquely mapped to the reference were used for differential expression analysis. All the files were grouped into 8 different replicate sets. Differential gene expression analysis was done using Qseq with a criteria of Log foldchange greater than or equal to 2, RPKM normalization, student t test, confidence interval of 95% with a p-value of 0.05. The list of Differentially expressed contigs with log2 Fold changes and RPKM normalizations was exported to do further analysis.
Annotation of Differentially Expressed Contigs:
An in-house python script was used to extract sequences for the differentially expressed contigs using the reference transcriptome. The output fasta file generated with differentially expressed contigs with the respective sequences is used for the annotation. Differentially expressed contigs were annotated using a web tool, Mercator-Mapman. This web site uses 7 different data bases with a custom blast cutoff and assign the contigs into 30 different bins such as RNA-Transcription, Signalling, Stress related, Photosynthesis related, CHO metabolism etc.,
The reads from the individual genotypes were separately mapped to the meta-reference transcriptome using lasergene suite, Array star software. Reads were also mapped on to genotype specific references created for each of the three genotypes. When the two results were compared, some interesting observations were made. The analysis using the meta reference showed significantly lesser number of differentially expressed genes compared to the mapping analysis using the genotype specific reference. For instance, in C76-16, when the reads were mapped to the genotype specific reference, 11701 genes were found at 90% confidence interval. But when same reads were mapped to the meta-reference, only 2698 genes were found at 90% confidence interval which is a significant reduction. This reduction was caused due to loss of genotype specific isoforms in the meta-reference created by merging the reads from all the three genotypes.
As the idea of a combined transcriptome did not yield the desired results, an alternate approach shall be used to address this problem. Instead of mapping the reads to a combined transcriptome, reads from individual genotypes will be mapped on to a individual reference specific to that genotype. Once the mapping is done and differentially expressed genes in each genotype are identified, they are annotated and the annotated gene lists across the three genotypes are compared to understand the differences in acclimation responses of these genotypes compared to each other.