Sporobolus stapfianus: Insights into desiccation tolerance in the resurrection grasses from linking transcriptomics to metabolomics

Abstract

Background: Understanding the response of resurrection angiosperms to dehydration and rehydration is critical for deciphering the mechanisms of how plants cope with the rigors of water loss from their vegetative tissues. We have focused our studies on the C₄ resurrection grass, Sporobolus stapfianus Gandoger, as a member of a group of important forage grasses.

Methods: We have combined non-targeted metabolomics with transcriptomics, via a NimbleGen array platform, to develop an understanding of how gene expression and metabolite profiles can be linked to generate a more detailed mechanistic appreciation of the cellular response to both desiccation and rehydration.

Results: The rehydration transcriptome and metabolome are primarily geared towards the rapid return of photosynthesis, energy metabolism, protein turnover, and protein synthesis during the rehydration phase. However, there are some metabolites associated with ROS protection that remain elevated during rehydration, most notably the tocopherols. The analysis of the dehydration transcriptome reveals a strong concordance between transcript abundance and the associated metabolite abundance reported earlier, but only in responses that are directly related to cellular protection during dehydration: carbohydrate metabolism and redox homeostasis. The transcriptome response also provides strong support for the involvement of cellular protection processes as exemplified by the increases in the abundance of transcripts encoding late embryogenesis abundant (LEA) proteins, anti-oxidant enzymes, early light-induced proteins (ELIP) proteins, and cell-wall modification enzymes. There is little concordance between transcript and metabolite abundance for processes such as amino acid metabolism that do not appear to contribute directly to cellular protection, but are nonetheless important for the desiccation tolerant phenotype of S. stapfianus.

Conclusions: The transcriptomes of both dehydration and rehydration offer insight into the complexity of the regulation of responses to these processes that involve complex signaling pathways and associated transcription factors. ABA appears to be important in the control of gene expression in both the latter stages of the dehydration and the early stages of rehydration. These findings add to the growing body of information detailing how plants tolerate and survive the severe cellular perturbations of dehydration, desiccation, and rehydration.

Keywords: Abiotic stress; Dehydration stress; Gene expression; Rehydration; Resurrection plants; Sporobolus stapfianus.

Fig. 1

Gene Ontology (GO) terms associated with significantly differentially abundant transcripts (SDATs) surveyed during the dehydration of young leaf tissues of S. stapfianus. Biological Network Gene Ontology (BiNGO) was used to determine biological process terms in the full GO terms (a) or the GO Slim plant terms (b) that were enriched (p < 0.05). Node size represents the number of genes within the node and the color filling represents the p-value (at p < 0.05), the darker the shade the lower the p value

Fig. 2

A heat map of SDATs encoding REDOX homeostasis related transcripts young leaf tissues of S. stapfianus. Multi Experiment Viewer (MeV 4.8.1) was used to generate the clustering of the data based on fold change values (log2). The columns represents the ratio between dehydrated (80%, 60, 40, and 30% RWC) as well as the dry state [DRY] and the hydrated state [HYD] for the first 5 columns and between either 12 and 24 h rehydration and the dry state [DRY] or initial hydrated states [HYD]. Red shading indicates a positive value for the fold change and green shading indicates negative values for the fold change in transcript abundance. Black indicates no change in transcript abundance

Fig. 3

A heat map of SDATs encoding energy metabolism related transcripts young leaf tissues of S. stapfianus. Multi Experiment Viewer (MeV 4.8.1) was used to generate the clustering of the data based on fold change values (log2). The color code represents the ratio between dehydrated (80%, 60, 40, and 30% RWC) as well as the dry state [DRY] and the hydrated state [HYD] for the first 5 columns and between either 12 and 24 h rehydration and the dry state [DRY] or initial hydrated states [HYD]. Red shading indicates a positive value for the fold change and green shading indicates negative values for the fold change in transcript abundance. Black indicates no change in transcript abundance

Fig. 4

Categorization of SDATs encoding members of transcription factor transcripts representing individual transcription factor families based on their response to dehydration in young leaf tissues of S. stapfianus. Dark and light gray shading indicate the counts of TFs per families that increased or decreased in abundance during dehydration, respectively

Fig. 5

A linear regression graph between transcript abundance derived from the qRT-PCR experiment (X-axis) and transcript abundance calculated from microarray analysis (Y-axis). Here, R² = 0.843, and the non-parametric Spearman correlation coefficient is 0.901. Each symbol represents the log2 ratio of each average of the conditions 80%, 60, 40, 30% RWC with respect to the average hydrated state [HYD] expression value for each of the 10 target genes described in Additional file 9: Table 7

Fig. 6

Endogenous ABA concentrations on a dry weight basis in young leaf tissues of S. stapfianus during dehydration (A) and rehydration (B). Analysis was conducted on hydrated tissues (HYD), tissues dehydrated to 60%, 40%, and 30% RWC as well as dry tissues [DRY]. Data are the means ± SD (n = 8)

Fig. 7

Concordance between the abundance of intermediates in the biosynthesis of Raffinose Family Oligosaccharides (RFO) and the transcripts encoding the enzymes of the pathway during the dehydration-rehydration cycle in young leaf tissues of S. stapfianus. A single longitudinal bar directly under the compound in the pathway represents the metabolite abundance heat map for each compound. The data used to generate the metabolite heat maps are reported in Oliver et al 2011. Red shading indicates a statistically significant fold increase in abundance and green shading a statistically significant fold decrease in metabolite abundance. White indicates no change in metabolite abundance. For the dehydrating samples (60% RWC to DRY) the fold change is given in relation to the hydrated control, where red shading indicates a statistically significant increase in abundance compared to the hydrated control for the dehydrating samples. For the rehydrating samples the fold change is given in relation to either the dry samples (/D) or the hydrated control (/H). The longitudinal bars directly associated with the enzyme identity in the pathway represent the transcript abundance heat map for each enzyme. Each line in the heat map constitutes a single contig annotated as encoding the associated enzyme. Red shading indicates a statistically significant the log₂-fold increase in abundance and green shading a statistically significant the log₂-fold decrease in abundance. Black indicates no change in transcript abundance. For the dehydrating samples (80% RWC to DRY) the log₂-fold change in abundance is given in relation to the hydrated control. For the rehydrating samples log₂-fold change in abundance is given in relation to either the dry samples (/DRY) or the hydrated control (/HYD)