Candidate regulators and target genes of drought stress in needles and roots of Norway spruce

Abstract

Drought stress impacts seedling establishment, survival and whole-plant productivity. Molecular responses to drought stress have been most extensively studied in herbaceous species, mostly considering only aboveground tissues. Coniferous tree species dominate boreal forests, which are predicted to be exposed to more frequent and acute drought as a result of ongoing climate change. The associated impact at all stages of the forest tree life cycle is expected to have large-scale ecological and economic impacts. However, the molecular response to drought has not been comprehensively profiled for coniferous species. We assayed the physiological and transcriptional response of Picea abies (L.) H. Karst seedling needles and roots after exposure to mild and severe drought. Shoots and needles showed an extensive reversible plasticity for physiological measures indicative of drought-response mechanisms, including changes in stomatal conductance (gs), shoot water potential and abscisic acid (ABA). In both tissues, the most commonly observed expression profiles in response to drought were highly correlated with the ABA levels. Still, root and needle transcriptional responses contrasted, with extensive root-specific down-regulation of growth. Comparison between previously characterized Arabidopsis thaliana L. drought-response genes and P. abies revealed both conservation and divergence of transcriptional response to drought. In P. abies, transcription factors belonging to the ABA responsive element(ABRE) binding/ABRE binding factors ABA-dependent pathway had a more limited role. These results highlight the importance of profiling both above- and belowground tissues, and provide a comprehensive framework to advance the understanding of the drought response of P. abies. The results demonstrate that a short-term, severe drought induces severe physiological responses coupled to extensive transcriptome modulation and highlight the susceptibility of Norway spruce seedlings to such drought events.

Keywords: Picea abies; ABA; Norway spruce; RNA-Seq; drought stress; gene expression; transcriptome.

Figure 1.

Physiological response of Norway spruce seedlings to drought. Norway spruce seedlings were subjected to increasing levels of water deficit (filled circles) or kept in well-watered control conditions (clear triangles). (a) Colored bars above the plot represent the following treatment conditions: soil moisture dropped from 80% (dark blue portion of the colored bar) to 30% FC after withholding water for 5 days and was regarded as mild water deficit (green portion of the colored bar). After maintaining this condition for 7 days (orange portion of the colored bar), a more severe drought stress was induced (red portion of the colored bar), again by withholding water. After 21 days, water-stressed plants were re-irrigated (cyan portion of the colored bar) and the soil moisture levels recovered to the level of well-watered plants. Colors indicating these experimental conditions are used in subsequent figures. All plants were visually identical, with no visible effects of drought for the sampled time points. An additional set of plants (n = 5) were not re-irrigated after the severe drought stress. All needles on these plants died after 23 days, indicated by a black cross. (b) Needle stomatal conductance (g_s) in control (clear triangles) and drought-stressed (filled circles) seedlings. (c) Abscisic acid levels of needles (green) and roots (brown) of drought-stressed seedlings. (d) Midday water potential (ψ_shoot) measured on shoots of three to five independent plants per time point in drought stress. The result of a one-way ANOVA performed on the average values of three measurements per plant is indicated. For all panels, data are means and error bars represent the ±95% CIs. Letter represent statistically significant differences (P < 0.05; Tukey’s HSD) between time points of the drought stress.

Figure 2.

Effect of drought on the transcriptomes of needles and roots in Norway spruce. Principal component analysis plot of transcriptomic data of drought-stressed needle and root (a) samples. Expression data of control, mild and severe drought-stressed seedlings and after re-irrigation were normalized using variance stabilizing transformation before ordination analysis. Three to four plants were sampled at each time point and were used for RNA-Seq. The first two components of the PCA are shown with samples colored by soil water percentage FC. (b) Venn diagrams of DEGs in needles (left) and roots (right) after comparing the mild and severe drought stages and after re-irrigation separately against well-watered expression levels. Independently, the number of DEGs and the number of these in common between the tissues are displayed separately for each stage of the treatment. Conditions are colored as in Figure 1c). Genes either up- or down-regulated in needles and roots and with the same expression in the tissues (common) are displayed in a bar graph and colored by treatment according to Figure 1.

Figure 3.

Gene Ontology Slim enrichment analysis of biological processes up-regulated in response to drought stress. Heatmaps of DEGs with expression over the eight sampling time points (Days 0, 2, 4, 5, 13, 18, 21 and 25 and colored as in Figure 1), separated by tissue, needles (green) and roots (brown). (a) Up-regulated genes of needle- (Clusters 1 and 2) and root-specific (Clusters 4–7) clusters or in common between the tissues (Cluster 3). Displayed are the Variance Stabilizing Transformation (vst) values scaled by row means. The seven highly populated clusters (Cl.) are detailed on the side, and the average trend of gene expression and a 95% CI are indicated for both tissues. Colored bars to the right of the heatmap correspond to the clusters represented in the line graphs as indicated by a colored bar to the right of each cluster line graph. (b) The lower panel of the figure lists all GOSlim biological processes categories represented in the clusters shown in (a), with columns colored by cluster. Bold text indicates categories that are significantly enriched. Circle sizes are relative to the category containing the largest number of genes; ‘response to stress’ (76 up-regulated genes). Circles in red represent significantly enriched categories (false discovery rate-adjusted P-value <0.05).

Figure 4.

Gene Ontology Slim enrichment analysis of biological processes down-regulated in response to drought stress. Heatmaps of DEGs with expression over the eight sampling time points (Days 0, 2, 4, 5, 13, 18, 21 and 25 and colored as in Figure 1), separated by tissue, needles (green) and roots (brown). (a) Down-regulated genes of needle- (Cluster 8) and root-specific (Clusters 10--12) Clusters or common between the tissues (Clusters 9 and 13). Displayed are the vst values scaled by row means. The six highly populated clusters (Cl.) are detailed on the side, and the average trend of gene expression and a 95% CI are indicated for both tissues. (b) The lower panel of the figure lists all GOSlim biological processes categories represented in the clusters shown in (a), with columns colored by cluster. Bold text indicates categories that were significantly enriched. Circle sizes are relative to the category containing the largest number of genes; ‘anatomical structure development’ (220 down-regulated genes). Circles in red represent significantly enriched categories (false discovery rate-adjusted P-value <0.05).

Figure 5.

Heatmap diagram of TFs differentially expressed in Norway spruce in response to drought. Displayed are the differentially expressed TFs over the eight sampling time points (Days 0, 2, 4, 5, 13, 18, 21 and 25 and colored as in Figure 1) and are separated by tissue, needles (green) and roots (brown). (a) Up-regulated TFs were clustered (Cl.) by expression and sorted by the same color coding as in Figure 3a and b). Down-regulated TFs were clustered by expression and sorted by the same color coding as in Figure 4a. Displayed are the vst values scaled by row means. In addition, information on sequence homology (Homol.) is visualized on the right side, with genes orthologous to Arabidopsis colored dark green, homologous to Arabidopsis in light blue, homologous to other gymnosperms in light green and Norway spruce-specific singletons in dark blue. Furthermore, TF families significantly enriched: ERFs, MYB-related, C2H2s and NACs in needle samples, and ERFs, MYBs and SAPs in roots are color-coded.

Figure 6.

Overview representation of physiological and gene expression responses to drought in Norway spruce seedlings. The left panel indicates representative profiles indicating changes in physiological indicators of drought stress. Profiles are representative and based on the data presented in Figure 1 (for stomatal conductance and ABA) or gene expression values (for root development and stress). On the right, the co-expression network is indicated with nodes (genes) colored to indicate their assigned network cluster. For three clusters, associated with ABA, root development or stress response (indicated by cluster gene ontology biological process enrichment results), expression profile plots are shown in which green profiles represent the expression in needles and brown represent the expression in roots.