Abstract
Plants encounter and respond to numerous abiotic stresses during their lifetimes. These stresses are often related and could therefore elicit related responses. There are, however, relatively few detailed comparisons between multiple different stresses at the molecular level. Here, we investigated the phenotypic and transcriptomic response of cultivated sunflower (Helianthus annuus L.) seedlings to three water-related stresses (i.e., dry-down, an osmotic challenge with polyethylene glycol 6000 [PEG], and salt stress), as well as a generalized low-nutrient stress. Our goal was to identify commonalities in the response to the three water-related stresses and compare them to a distinct low-nutrient stress. All four stresses negatively impacted seedling growth, with the low-nutrient stress having a more divergent response from control as compared to the water-related stresses. Observed phenotypic responses were consistent with expectation for growth in low-resource environments, including increased (i.e., less negative) carbon fractionation values and leaf C:N ratios, as well as increased belowground biomass allocation. Analysis of the leaf and root transcriptome under each stress scenario revealed that most genes were differentially expressed in response to multiple stresses. The number of differentially expressed genes (DEGs) under stress was greater in leaf tissue, but roots exhibited a higher proportion of DEGs unique to individual stresses. Overall, the three water-related stresses had a more similar transcriptomic response to each other vs. low-nutrient stress, though this pattern was more pronounced in root tissue than in leaf tissue. In contrast with the results of our differential expression analysis, co-expression network analysis revealed that the response to each of the four stresses in our study were generally non-overlapping and there was little indication of a shared co-expression response despite the majority of DEGs being shared between multiple stresses. Importantly, PEG stress, which is often used to simulate drought stress in experimental settings, had little transcriptomic resemblance to true water limitation (i.e., dry-down) in our study calling into question its utility as a means for simulating drought.