Insights into the Role of Transcriptional Gene Silencing in Response to Herbicide-Treatments in Arabidopsis thaliana

Abstract

Herbicide resistance is broadly recognized as the adaptive evolution of weed populations to the intense selection pressure imposed by the herbicide applications. Here, we tested whether transcriptional gene silencing (TGS) and RNA-directed DNA Methylation (RdDM) pathways modulate resistance to commonly applied herbicides. Using Arabidopsis thaliana wild-type plants exposed to sublethal doses of glyphosate, imazethapyr, and 2,4-D, we found a partial loss of TGS and increased susceptibility to herbicides in six out of 11 tested TGS/RdDM mutants. Mutation in REPRESSOR OF SILENCING 1 (ROS1), that plays an important role in DNA demethylation, leading to strongly increased susceptibility to all applied herbicides, and imazethapyr in particular. Transcriptomic analysis of the imazethapyr-treated wild type and ros1 plants revealed a relation of the herbicide upregulated genes to chemical stimulus, secondary metabolism, stress condition, flavonoid biosynthesis, and epigenetic processes. Hypersensitivity to imazethapyr of the flavonoid biosynthesis component TRANSPARENT TESTA 4 (TT4) mutant plants strongly suggests that ROS1-dependent accumulation of flavonoids is an important mechanism for herbicide stress response in A. thaliana. In summary, our study shows that herbicide treatment affects transcriptional gene silencing pathways and that misregulation of these pathways makes Arabidopsis plants more sensitive to herbicide treatment.

Keywords: 2,4-D; ROS1; chromatin mutants; epigenetics; glyphosate; herbicide resistance; imazethapyr.

Figure 1

Herbicide stress effect on epigenetically regulated transcription. (a) Representative β-Glucuronidase (GUS)-stained shoots of the plants of line L5, 48 h after treatment with glyphosate, imazethapyr (IM) and 2,4-D, non-treated plant (L5 control) and L5 ddm1 plant serving as a positive control. Plants line L5 are carrying a single insert of a multicopy 35S Cauliflower Mosaic Virus Promoter driven GUS (Pro35S::GUS) locus in L5/6b5 line, gene suppressed by transcriptional gene silencing (TGS), all mutants are in the Col-0 background. Scale bar = 1 cm. (b) Relative transcript abundance of GUS in L5 measured by RT-qPCR 48 h after the treatment. Error bars indicate the standard deviation of the three biological replicates. Statistically significant differences relative to L5 control are indicated by asterisks (t-test, p ≤ 0.05); * compared to L5 Control.

Figure 2

(a) Effect of sublethal doses of glyphosate, imazethapyr and 2,4-D on Arabidopsis thaliana wild type (WT) and selected chromatin mutants 10 days after treatment (DAT). (b–d) Shoot dry weight from treated plants compared to non-treated plants (% untreated check) after (b) glyphosate, (c) imazethapyr and (d) 2,4-D treatment. The mean of each study (1 and 2) was composed by three biological replicates, studies 1 and 2 followed by different letter differ significantly by Tukey’s test (p ≤ 0.05); ns = nonsignificant. Treated mutants followed by asterisk differ significantly from treated WT, according to Tukey’s test (p ≤ 0.05); * compared to WT.

Figure 3

(a) Visual effect (b) shoot dry weight of A. thaliana Col-0 wild type (WT) and ros1 mutant corresponding to study 3, 10 days after treatment (DAT) of glyphosate and imazethapyr (IM). Error bars indicate the standard deviation of four biological replicates, statistically significant differences relative to WT is indicated by asterisk (t-test, p ≤ 0.05); * compared to WT.

Figure 4

HPLC measurement of global 5-mdC (%) in wild type (WT) and ros1, in the control condition and after treatment with imazethapyr (IM), 48 h after treatment. ddm1 and lambda phage DNA were used as low and high DNA methylation controls, respectively. Error bars denote standard deviations from three biological replicates. Means followed by a different letter differ significantly according to Tukey’s test (p ≤ 0.05).

Figure 5

(a) Differentially expressed genes (DEGs) in response to imazethapyr (IM) and in the control condition in wild type (WT) and ros1 mutant. The number of up and downregulated genes is represented in black and gray bars, respectively. The differences in gene expression were obtained based on the Log2 Fold Change ≥2 and adjusted p-value < 0.05 in Cuffdiff. (b) Venn diagrams showing the overlap of genes comparing genes inducted by IM in WT (herbicide effect—blue circle), with repressed genes in ros1 (mutant effect—yellow circle) and effect of IM in ros1 (herbicide effect on mutant—green circle).

Figure 6

Detailed information of biological process representing the percent of genes involved in biological process pathways of 89 genes of interest induced by imazethapyr (IM) in wild type (WT) and downregulated in ros1 mutant, performed by using singular enrichment analysis (SEA) in AgriGo. Gray and black bars indicate the percent of genes related to the input list and the percent of genes compared to genome reference, respectively.

Figure 7

Transcriptionally regulated genes in response to herbicide application in wild type (WT) A. thaliana are not upregulated in the ros1 mutant. (a) Quantitative PCR validation (fold-change) of genes from the differentially expressed genes (DEGs) profiling. Error bars indicate the standard deviation of three biological replicates and asterisks indicate significant differences between the treatments and A. thaliana wild type (WT) control according to t-test (p < 0.05); * compared to WT control. (b) Differentially expressed genes (DEGs) were performed by using CummeRbund, comparing genes in response to imazethapyr (IM) in A. thaliana wild type (WT) and ros1, the gene-normalized signal intensities are shown in the heat map.

Figure 8

TT4 function for the rate-limiting step into flavonol production is necessary for A. thaliana tolerance to low herbicide dose. (a) Visual effect, (b) shoot dry weight of A. thaliana wild type (WT) and tt4 mutant 14 days after application of imazethapyr (IM). Error bars indicate the standard deviation of four biological replicates, statistically significant differences relative to WT is indicated by asterisk (t-test, p ≤ 0.05); * compared to WT.