Transcriptome differences between two sister desert poplar species under salt stress

Abstract

Background: Populus euphratica Oliv and P. pruinosa Schrenk (Salicaceae) both grow in dry desert areas with high summer temperatures. However, P. euphratica is distributed in dry deserts with deep underground water whereas P. pruinosa occurs in deserts in which there is underground water close to the surface. We therefore hypothesized that these two sister species may have evolved divergent regulatory and metabolic pathways during their interaction with different salt habitats and other stresses. To test this hypothesis, we compared transcriptomes from callus exposed to 24 h of salt stress and control callus samples from both species and identified differentially expressed genes (DEGs) and alternative splicing (AS) events that had occurred under salt stress.

Results: A total of 36,144 transcripts were identified and 1430 genes were found to be differentially expressed in at least one species in response to salt stress. Of these DEGs, 884 and 860 were identified in P. euphratica and P. pruinosa, respectively, while 314 DEGs were common to both species. On the basis of parametric analysis of gene set enrichment, GO enrichment in P. euphratica was found to be significantly different from that in P. pruinosa. Numerous genes involved in hormone biosynthesis, transporters and transcription factors showed clear differences between the two species in response to salt stress. We also identified 26,560 AS events which were mapped to 8380 poplar genomic loci from four libraries. GO enrichments for genes undergoing AS events in P. euphratica differed significantly from those in P. pruinosa.

Conclusions: A number of salt-responsive genes in both P. euphratica and P. pruinosa were identified and candidate genes with potential roles in the salinity adaptation were proposed. Transcriptome comparisons of two sister desert poplar species under salt stress suggest that these two species may have developed different genetic pathways in order to adapt to different desert salt habitats. The DEGs that were found to be common to both species under salt stress may be especially important for future genetic improvement of cultivated poplars or other crops through transgenic approaches in order to increase tolerance of saline soil conditions.

Figure 1

Venn diagrams showing mapped genes expressed in each possible pair out of the four libraries. PeuC, P. euphratica control callus; PeuS, P. euphratica salt-stressed callus; PprC, P. pruinosa control callus; PprS, P. pruinosa salt-stressed callus.

Figure 2

Comparison of four metrics for classifying DEGs. Venn diagrams of the numbers of up-regulated (left) and down-regulated (right) genes identified by four comparisons of control callus and salt-stressed callus from P. euphratica (top) and P. pruinosa (bottom).

Figure 3

Number of DEGs in P. euphratica and P. pruinosa . The numbers of DEGs that were exclusively up- or down-regulated in one species are shown in each circle. The numbers of DEGs with the same or opposite pattern of expression changes between the two species are shown in the overlapping regions. The total numbers of up- or down-regulated genes in each species are shown outside the circles.

Figure 4

Expression pattern validation of selected genes by qRT-PCR. Expression changes of 21 DEGs in the salt-stressed calli relative to the control calli were measured by qRT-PCR. The transcriptional level of candidate genes was examined by real time PCR with three biological replications and actin was used as an internal control. Results were present as target/reference ratios normalized by the calibrator. No significant differences were shown between qRT-PCR and the Illumina data (Pearson’s correlation coefficient r = 0.8). The Y-axis indicates the fold change of transcript abundance in salt-stressed callus relative to the control callus. PeuC, P. euphratica control calli; PeuS, P. euphratica salt-stressed calli; PprC, P. pruinosa control calli; PprS, P. pruinosa salt-stressed calli.

Figure 5

Gene ontology (GO) annotation of salt-responsive genes compared between P. euphratica and P. pruinosa. WEGO was used to produce the graph. We divided the sets into the three major GO domains: biological process, cellular component and molecular function, and the number (right y-axis) and percentage (left y-axis) of genes were calculated.

Figure 6

Hierarchical clustering of 58 genes related to plant hormone biosynthesis. Hierarchical cluster analysis was conducted using the software PermutMatrix v1.9.3 and displayed as a log-(FPKM) value for the intensity of expression of each DEG. FPKM, number of fragments per kilobase of exon per million fragments mapped; PeuC, P. euphratica control callus; PeuS, P. euphratica salt-stressed callus; PprC, P. pruinosa control callus; PprS, P. pruinosa salt-stressed callus.

Figure 7

Number of loci showing AS events in P. euphratica and P. pruinosa . The numbers of loci undergoing AS events in each species and treatment are shown. PeuC, P. euphratica control callus; PeuS, P. euphratica salt-stressed callus; PprC, P. pruinosa control callus; PprS, P. pruinosa salt-stressed callus.