Transcriptional signatures associated with waterlogging stress responses and aerenchyma formation in barley root tissue

Abstract

Background and aims: The frequency of extreme precipitation events is predicted to increase owing to climate change, leading to soil waterlogging and crop yield losses, particularly in the case of susceptible species, such as barley (Hordeum vulgare). Aerenchyma formation is a key morphological adaptation to waterlogging stress and hypoxic conditions; however, its genetic regulation in barley remains largely unresolved. The aim of this study was to address this knowledge gap and characterize the transcriptional signatures associated with the waterlogging stress response and aerenchyma formation in barley roots.

Methods: Two barley cultivars (Franklin and Yerong) were subjected to waterlogging stress, followed by analysis of phenotypic traits, including root aerenchyma formation, and transcriptomic profiling of root tissue. Differential gene expression analysis and gene regulatory network construction were carried out using generated RNA-sequencing datasets.

Key results: Performed analyses identified genes transcriptionally responsive to 24 and 72 h of waterlogging in both cultivars and highlighted metabolic adaptations, regulation of reactive oxygen species signalling and management of stress responses as key elements of the waterlogging response in barley roots. Large intra-individual variation was observed for root aerenchyma formation. This variation was exploited to identify 81 candidate aerenchyma-associated genes and ascertain pathways involved in aerenchyma formation. Furthermore, network analyses suggested that the DNA damage response gene DRT100 and the cell wall-modifying genes XTH16 and XTH15 are regulatory hub genes in aerenchyma formation.

Conclusions: This study provides new insights into transcriptional signatures associated with waterlogging responses and aerenchyma formation in barley roots. The identified candidate aerenchyma-associated genes offer new targets for future research and breeding efforts aimed at enhancing waterlogging tolerance in this crop species.

Keywords: Barley; aerenchyma; flooding; root transcriptomics; waterlogging.

Fig. 1.

Time line for RNA-seq sampling. Barley plants were grown and sampled according to the presented time line: seeds were cold stratified for 3 days, transferred to 20 °C in the dark for germination for 4 days before planting in vermiculite soaked in 0.25 strength Hoagland's solution. Plants were grown under 16 h light–8 h dark, 20 °C constant-temperature conditions and watered three times a week with 0.25 strength Hoagland's solution from above, with excess removed after 30 min. This was carried out for 21 days, until plants reached the three-leaf stage and were subsequently waterlogged with nutrient solution in plastic bags. Control plants were watered three times a week according to the normal watering schedule. Franklin and Yerong cultivars were sampled at 24 and 72 h for RNA sequencing, and for aerenchyma analysis and phenotypic traits at 1, 3, 7 and 21 days of waterlogging treatment.

Fig. 2.

Response to waterlogging treatment observed in barley cultivars Franklin and Yerong. The waterlogging treatment was initiated at the three-leaf stage. (A) Shoot growth was determined by measuring plant height prior to start of the treatment and after 1, 3, 7 and 21 days of waterlogging treatment. (B) Aerenchyma percentage in adventitious roots. Roots (two per plant) were sampled 3–5 cm from the root tip, and log₁₀-transformed values were used for statistical analyses. (C) Mean total length of adventitious roots per plant. (D) Mean number of adventitious roots for each plant. Bars represent mean, error bars represent 95 % confidence interval. ***P ≤ 0.001, **P ≤ 0.01 and *P ≤ 0.05 after two-way ANOVA (P-value = 0.05) with Bonferroni correction for multiple testing using estimated marginal means in SPSS. Duration corresponds to the duration of treatment. Experiments were repeated three times, with n = 4 plants sampled per cultivar–treatment combination at days 1 and 3 of waterlogging, n = 2–3 plants at day 7, and n = 3–4 at day 21.

Fig. 3.

Observation of intra-individual variation in aerenchyma formation in adventitious roots. Following 1 and 3 days of treatment, root systems were washed in water, and two roots were sampled from each plant for RNA sequencing (segment 1–3 cm from the root tip) and aerenchyma percentage analysis (segment 3–5 cm from the root tip). For aerenchyma analysis, roots were embedded in liquid agar prior to hand-sectioning thin cross-sections of root tissue and imaging using a light microscope at ×100 magnification. Percentage aerenchyma was quantified for each root using ImageJ software. From each plant, both roots were compared and categorized as high (H_AE) or low (L_AE) aerenchyma-forming roots relative to one another, which led to the detection of high intra-individual aerenchyma percentage in roots sampled from the same plant (A, B). The graph indicates intra-individual variation in aerenchyma percentage of roots from the same plant, and the images represent an example of intra-individual variation in % AE between two roots from the same waterlogged plant (cultivar Yerong; scale bar: 200 µm). Duration corresponds to the duration of treatment. ***P ≤ 0.001 and **P ≤ 0.01 based on two-way repeated-measures ANOVA (P-value = 0.05) with Bonferroni correction for multiple testing using estimated marginal means; error bars display 95 % confidence interval. Experiments were repeated three times; each time, n = 4 plants were analysed for each cultivar, treatment, time point and aerenchyma group.

Fig. 4.

Transcriptional response to waterlogging stress in barley roots. (A) UpSet plots representing the overlapping DEGs upregulated (red) and downregulated (blue) in response to waterlogging (WL) at 24 and 72 h treatment in comparison to the respective control in each cultivar. Gene list size represents the number of DEGs for each comparison, and genes per intersection corresponds to the number of shared DEGs among the comparisons marked as connected dots below. The single dots represent DEGs unique to one comparison. Each comparison represents DEGs under waterlogging treatment with control as the baseline level. (B) Venn diagram representing the general waterlogging responses. General waterlogging response genes are those commonly differentially regulated at point time points and in both cultivars. (C, D) The number of DEGs unique to the 24 h (C) and 72 h (D) time points are shown in respective Venn diagrams indicating DEGs common to both cultivars and uniquely differentially expressed. Upregulated genes are shown in red and downregulated genes in blue. Franklin unique expressed genes are highlighted in purple areas; Yerong unique genes are shown in yellow areas.

Fig. 5.

Gene ontology (GO) enrichment of WL response genes. Upregulated and downregulated DEGs associated with core (both cultivars), for general (both time points) (A), 24 h unique (B) and 72 h unique (C) time points were identified following DESeq2 analysis. BaRTv2 gene names were converted to MorexV3 gene names (represented by n for each GO analysis) and subsequently input into ShinyGO v.0.75c. ShinyGO output was set to 30 GO terms and FDR cut-off as 0.05. The lollipop graph shows fold enrichment on the x-axis with the size of dot representing number of genes associated with a particular GO term. Colour of line and dot represents −log₁₀(FDR).

Fig. 6.

Harnessing intra-individual variability in aerenchyma formation to identify aerenchyma-associated transcriptomic response in Franklin and Yerong barley cultivars. (A) Differential gene expression analysis was carried out for high percentage aerenchyma (H_AE) versus low percentage aerenchyma (L_AE) comparisons in both treatments (waterlogging or control), both cultivars (Franklin or Yerong) and both time points (24 h or 72 h), with results summarised in the UpSet plot. Gene list size represents the number of DEGs for each sample for each comparison, and genes per intersection represents the number of overlapping DEGs between specific comparisons as indicated by the connected dots below. Single dots represent unique DEGs to one comparison. Each comparison represents DEGs in high percentage aerenchyma forming sample with low percentage aerenchyma sample as the baseline level. Red box indicates aerenchyma-associated DEGs present in at least two of eight of the high versus low percentage aerenchyma comparison groups (81 DEGs total). (B, C) GO enrichment of aerenchyma-associated DEGs. Aerenchyma-associated DEGs (BaRTv2 gene format) were converted to MorexV3 gene name where possible, with n indicating the number of genes converted to MorexV3 format. (B) MorexV3 gene names were subsequently input ShinyGO v.0.75c for GO enrichment. Output was set to 30 GO terms and FDR cut-off as 0.05. (C) Aerenchyma-associated DEGs were input as BaRTv2 protein FASTA sequences into the ID mapping tool (BLASTp) in PlantRegMap to obtain Arabidopsis thaliana IDs; n represents number of DEGs which had a corresponding A. thaliana ID. The A. thaliana IDs were input into the GO term enrichment tool in PlantRegMap, with P-value cut-off of 0.01. (D) Clustered gene regulatory network (GRN) of aerenchyma-associated genes with predicted hub genes. Aerenchyma-associated genes (dark purple) with corresponding Arabidopsis gene IDs (n = 33) were input into Cytoscape using the GeneMANIA v.3.5.2 plugin, where 20 resulting (light purple) nodes were inserted based predicted gene interactions. Node size corresponds to the degree of connectivity of nodes, with larger nodes being more connected. Hub genes are identified by nodes with the greatest connectivity and circled in green. The GeneMANIA network was clustered using the Cytoscape plugin clusterMaker2 v.2.3.4 with MCL clustering, granularity parameter set to 4. Interaction links in purple are predicted based on gene co-expression, in blue are co-localized genes, and yellow are genes with shared protein domains. The edge thickness in the GRN corresponds to normalized maximum weight.

Fig. 7.

Summary of transcriptional changes in barley roots under waterlogging stress.