Herboxidiene triggers splicing repression and abiotic stress responses in plants

Abstract

Background: Constitutive and alternative splicing of pre-mRNAs from multiexonic genes controls the diversity of the proteome; these precisely regulated processes also fine-tune responses to cues related to growth, development, and stresses. Small-molecule inhibitors that perturb splicing provide invaluable tools for use as chemical probes to uncover the molecular underpinnings of splicing regulation and as potential anticancer compounds.

Results: Here, we show that herboxidiene (GEX1A) inhibits both constitutive and alternative splicing. Moreover, GEX1A activates genome-wide transcriptional patterns involved in abiotic stress responses in plants. GEX1A treatment -activated ABA-inducible promoters, and led to stomatal closure. Interestingly, GEX1A and pladienolide B (PB) elicited similar cellular changes, including alterations in the patterns of transcription and splicing, suggesting that these compounds might target the same spliceosome complex in plant cells.

Conclusions: Our study establishes GEX1A as a potent splicing inhibitor in plants that can be used to probe the assembly, dynamics, and molecular functions of the spliceosome and to study the interplay between splicing stress and abiotic stresses, as well as having potential biotechnological applications.

Keywords: ABA; Abiotic stress responses; Alternative splicing; GEX1A; Pladienolide B; SR proteins; Splicing inhibitors.

Fig. 1

GEX1A inhibited plants growth and development. a, the structure of GEX1A (herboxidiene). b, Effects of GEX1A on Arabidopsis seed germination. GEX1A inhibits seed germination in a dose-dependent manner. c, Inhibition of primary root elongation of Arabidopsis seedlings by GEX1A. 5-day-old Col-0 seedlings transferred from MS medium containing DMSO (control), 0.2 μM, 0.5 μM, and 1 μM GEX1A for an additional 4 days. d, GEX1A inhibits tomato seeds germination. Tomato seeds was incubated in water with DMSO (negative control), 1 μM and 5 μM GEX1A for 8 days. e, GEX1A inhibits rice root elongation. The rice seeds were germinated on ½ MS plate for 3 days, then transfer onto ½ MS with 1 μM GEX1A for 2 days. The red bar marks root tip of the transferring time. f, comparison inhibition effect of PB and GEX1A on Arabidopsis root growth, “√” stands for 0% root elongation rate on the chemical

Fig. 2

GEX1A alters the splicing patterns of a set of genes. The cDNAs were prepared from one-week-old Arabidopsis seedlings treated with GEX1A (5 μM) for 6 h, with DMSO as control. RT-PCR was performed using primers that flank introns, the gene name/locus identifier is shown. D6, DMSO 6 h, G6, GEX1A 6 h

Fig. 3

Gene expression changed by GEX1A corresponding to stress responses. a, upper panel, a heatmap was generated by mapping 400 randomly chosen upregulated genes in GEX1A treatment to the microarray database using Genevestigator. The heatmap indicates that a great number of these genes are upregulated (red) by ABA, drought and salt stress. Bottom panel: a heatmap was generated by mapping 400 randomly chosen down-regulated genes in the GEX1A treatment to the microarray database using Genevestigator. The heatmap indicates that a great number of these genes are downregulated (green) by ABA, drought, and salt stress. b, Functional categorization of regulated genes. Functional categorization (biological process) of up/downregulated genes in GEX1A treatment. Top 25 enriched pathways are shown. c, Differentially expressed genes in GEX1A treatment were mapped onto the response-to-abscisic-acid pathway. The analysis was performed using the Exploratory Gene Association Networks (EGAN) software tool. Orange lines show the participation of the genes in abscisic acid-activated signaling pathways and blue lines show known interactions between the genes connected. Green ovals represent upregulated genes and blue ovals represent downregulated genes

Fig. 4

Comparison of differentially expressed genes in response to GEX1A and PB. a, Venn diagram showing a comparison of the upregulated genes identified in 6 h GEX1A treatment and 6 h PB treatment. b, Venn diagram showing a comparison of the downregulated genes identified in 6 h GEX1A treatment and 6 h PB treatment. c, Functional categorization (biological process) of upregulated genes in both GEX1A and PB treatments. The top 20 enriched pathways are shown. d, Functional categorization (biological process) of upregulated genes in only GEX1A treatment. The top 20 enriched pathways are shown

Fig. 5

GEX1A treatment induced RD29A-LUC expression and stomatal aperture closure, caused relocation of SR45 proteins. a, One-week-old RD29A-LUC transgenic seedlings were treated with DMSO, 100 μM ABA or 5 μM GEX1A for 6 h, then sprayed with D-luciferase and observed by CCD camera. b, Relative bioluminescence intensities of RD29A-LUC seedlings in each treatment. c, Leaves of 3–4-week-old Arabidopsis plants were treated in opening solution for approximately 2.5 h and then transferred into opening solution with 20 μM GEX1A for 4 h. DMSO and ABA were used as negative and positive controls, respectively. d, Boxplot comparison of stomatal aperture in Arabidopsis leaves after the indicated treatments, three replicates and 150 stomata were measured. e, One-week-old 35S::SR45:GFP transgenic seedlings were treated with DMSO (left) or 5 μM GEX1A (right) for 24 h. Left-upper, GFP signal in the elongation zone of a 35S::SR45:GFP root in control conditions. Left-bottom, close-up view of nuclei of elongation zone cells from DMSO-treated 35S:SR45:GFP transgenic plants. Right-upper, GFP signal in the elongation zone of a 35S::SR45:GFP root in GEX1A treatment. Right-bottom, close-up view of nuclei of elongation zone cells from 5 μM GEX1A-treated 35S::SR45:GFP transgenic plants, nuclear speckles formed in the nuclei

Fig. 6

Genes with perturbed splicing in GEX1A treatment are associated with stress responses. a and b, Comparison of intron retention between control and GEX1A treatments. Reads numbers for the exons and introns are plotted. The expression of introns (a), but not exons (b), in GEX1A treatments showed a global upregulation. c, Comparison of global alternative splicing between control and GEX1A treatments. The intron retention events increased in the drug-treated samples, while the other AS events (including alternative 5’ splice sites, 3’ splice sites, and exon skipping) decreased in the GEX1A-treated samples. d, A two-dimensional view of the functional annotations of genes with retained introns in GEX1A treatment. The functional classification of genes was done using the DAVID software. The top 40 functional annotations were ordered by the number of genes in each category and selected for the two-dimensional view, which indicates that genes with retained introns were strikingly enriched in the response-to-abiotic-stress category. e, Functional category (biological process) of genes with retained introns in GEX1A treatment. The top 20 categories were ordered by the enrichment scores and selected. f, RT-PCR. The cDNAs were prepared from one-week-old Arabidopsis seedlings treated with 5 μM GEX1A for 6 h, DMSO as control. Gene structure and intron retention of interesting regions from eight genes were shown in IGV program, validation of intron retention of each gene was performed by RT-PCR using intron-flanking primers, with the result shown on the right. The red bar in the IGV program snapshot represents the target amplification region. D6, DMSO 6-h treatment; G6, GEX1A 6-h treatment

Fig. 7

Comparison of intron-retention events and genes between GEX1A and PB treatments. a, Venn diagram showing a comparison of the intron-retention events identified in 6-h GEX1A treatment and 6-h PB-treatment. b, Venn diagram showing a comparison of the intron retention genes identified in 6-h GEX1A treatment and 6-h PB treatment. c, Functional category (biological process) of 8039 genes with retained introns. Each of these genes has the same intron-retention events in both treatments. The top 20 categories were ordered by the enrichment scores and displayed. d, Functional category (biological process) of genes with retained introns only identified in 6-h GEX1A treatment, when compared with 6-h PB treatment. The top 20 categories were ordered by the enrichment scores and displayed

Fig. 8

Plants with reduced ABA sensitivity are partially resistant to GEX1A. a, The abi1-1C mutant is partially resistant to PB in seedling establishment compared to wild-type Col-0 plants. Quantification of seedling establishment (seedlings developing a first pair of true leaves) was performed on MS plates supplemented with DMSO (Control, white bars), 0.2 μM GEX1A (black bars) or 1 μM ABA (gray bars) 8 days after the seeds were sown. Values are averages of 3 independent experiments ± SD (n > 100). * indicates a p-value ≤ 0.05 by t-test compared to wild type under the same treatment. b, Photograph of representative seedlings from a. c, The abi1-1C mutant is partially resistant to GEX1A in seed germination. Seeds were stratified for 72 h in cold and seed germination (scored as radicle emergence) was calculated 48 h after transfer of the seeds to the controlled growth condition chamber. Values are average of three independent experiments ± SD (n > 100). * indicates a p-value ≤ 0.05 by t-test compared to wild type under the same treatment. Seeds were sown on MS plates supplemented with DMSO (Control, white bars), 1 μM GEX1A (black bars) and 1 μM ABA (gray bars). d, Photograph of representative seedlings showing sensitivities of mutants to ABA and GEX1A. e, Plants with reduced sensitivity to ABA are partially resistant to PB in root growth assay. Seedlings grown vertically on MS plates for 3 days were transferred to MS plates containing DMSO (Control, white bars), 0.2 μM GEX1A (black bars) or 10 μM ABA (gray bars). Root length was calculated with ImageJ 7 days after the transfer. Values are average of three independent experiments ± SD (n > 10). * indicates a p-value ≤ 0.05 by t-test compared to wild type under the same treatment. f, 5-day-old Arabidopsis Col (0) wild-type and sr45-1 mutant seedlings were transferred onto 1⁄2 MS medium with 0.2 μM GEX1A for 4 days. The position of the root tip of seedlings when they were transferred is shown by the black bars

Fig. 9

GEX1A affected the splicing of PP2C and SnRK2 genes differently. The cDNAs were prepared from one-week-old Arabidopsis seedlings treated with 5 μM GEX1A or 25 μM ABA for 6 h, with DMSO as control. RT-PCR was performed using primers flanking the first and last exon of each gene. ”D” indicates DMSO treatment, “G” indicates 5 μM GEX1A treatment and “A” indicates 25 μM ABA treatment; gene names are indicated at the bottom of each panel. Functional transcripts of most of PP2C genes were removed by strong intron retention in GEX1A-treated plants, whereas under the same conditions, SnRK2.2, SnRK2.3, and SnRK2.6 kept producing functional transcripts with varying levels of intron retention. ABA did not cause obvious intron retention in the selected genes, when compared with DMSO treatment