The RNA silencing enzyme RNA polymerase v is required for plant immunity

Abstract

RNA-directed DNA methylation (RdDM) is an epigenetic control mechanism driven by small interfering RNAs (siRNAs) that influence gene function. In plants, little is known of the involvement of the RdDM pathway in regulating traits related to immune responses. In a genetic screen designed to reveal factors regulating immunity in Arabidopsis thaliana, we identified NRPD2 as the OVEREXPRESSOR OF CATIONIC PEROXIDASE 1 (OCP1). NRPD2 encodes the second largest subunit of the plant-specific RNA Polymerases IV and V (Pol IV and Pol V), which are crucial for the RdDM pathway. The ocp1 and nrpd2 mutants showed increases in disease susceptibility when confronted with the necrotrophic fungal pathogens Botrytis cinerea and Plectosphaerella cucumerina. Studies were extended to other mutants affected in different steps of the RdDM pathway, such as nrpd1, nrpe1, ago4, drd1, rdr2, and drm1drm2 mutants. Our results indicate that all the mutants studied, with the exception of nrpd1, phenocopy the nrpd2 mutants; and they suggest that, while Pol V complex is required for plant immunity, Pol IV appears dispensable. Moreover, Pol V defective mutants, but not Pol IV mutants, show enhanced disease resistance towards the bacterial pathogen Pseudomonas syringae DC3000. Interestingly, salicylic acid (SA)-mediated defenses effective against PsDC3000 are enhanced in Pol V defective mutants, whereas jasmonic acid (JA)-mediated defenses that protect against fungi are reduced. Chromatin immunoprecipitation analysis revealed that, through differential histone modifications, SA-related defense genes are poised for enhanced activation in Pol V defective mutants and provide clues for understanding the regulation of gene priming during defense. Our results highlight the importance of epigenetic control as an additional layer of complexity in the regulation of plant immunity and point towards multiple components of the RdDM pathway being involved in plant immunity based on genetic evidence, but whether this is a direct or indirect effect on disease-related genes is unclear.

Figure 1. Characterization of ocp1 plants.

(A) Comparative histochemical analysis of GUS activity in rosette leaves from a parental wild-type plant carrying the P_Ep5C:GUS transgene (left), and ocp1 mutant plant (right). (B) Macroscopic comparison of 3-week-old wild-type (left) and ocp1 plants (right). (C) Resistance response of wild-type and ocp1 plants to virulent B. cinerea. Lesion size was measured 5 days after inoculation (dpi). Data points represent average lesion size ± SE (n≥30 lesions). Representative leaves from wild-type and ocp1 plants 4 dpi. (D–E) Growth rates of virulent PsDC3000 (D) and avirulent PsDC3000 (AvrRpm1) (E) in Col-0, ocp1 and npr1 or rpm1 plants. (F–G) RT-qPCR expression analysis of PR-1 (F) and PDF1.2a (G) in wild-type and ocp1 plants at different times following inoculation with PsDC3000 (F) and B. cinerea (G). Data represent the mean ± SD; n = 3 biological replicates.

Figure 2. ocp1 is a mutant allele of NRPD2.

(A) OCP1 corresponds to At2g27040 encoding NRPD2. The G nucleotide residue deleted in the ocp1 allele is indicated in red bold uppercase letters in the wild-type sequence. Deduced amino acid sequences are indicated below each nucleotide triplet, and the first amino acid change (S to T) where the frameshift of the OCP1 protein starts is shown in blue. (B) NRPD2 expression level by RT-PCR in mRNAs derived from Col-0, nrpd2-2 and ocp1 plants. The eEF1a house-keeping gene was used as a control. (C–D) nrpd2 plants show enhanced susceptibility to fungal pathogens. Lesion size was measured in Col-0, ocp1 and nrpd2-2 plants after inoculation with B. cinerea (C) or P. cucumerina (D). Data points represent average lesion size ± SE (n≥30 lesions). ANOVA detected significant differences at the P<0.05 level. (E–F) RT-qPCR determination of PDF1.2a (E) and PR-1 (F) transcript levels following inoculation with P. cucumerina. Data represent the mean ± SD; n = 3 biological replicates.

Figure 3. ocp1 plants show hypomethylation in RdDM target DNA sequences.

Genomic DNA isolated from Col-0, ocp1 and ago4-2/ocp11 plants (A) and nrpd2, nrpd1 and nrpe1 (B) was digested (+) or not (−) with HaeIII and amplified by PCR for SUPERMAN promoter (A), the ribosomal 5S genes and the retrotransposon AtSN1 (B). ABI5 contains no target sequences for HaeIII and was used as a control.

Figure 4. Comparative immune responses of RdDM mutants to inoculation with B. cinerea and P. cucumerina.

(A) Disease susceptibility of Col-0, nrpd1, nrpe1, nrpd2, ocp1 and ago4-2/ocp11 plants to B. cinerea. (B) Comparative disease susceptibility of the Pol IV and Pol V defective mutants to P. cucumerina. (C–D) RT-qPCR of PDF1.2a (C) and PR-1 (D) transcript levels following inoculation with P. cucumerina in Col-0, nrpd1, nrpe1 and nrpd2 plants. Data represent the mean ± SD; n = 3 biological replicates. (E) Comparative disease susceptibility of rdr2, drd1, drm1drm2 and nrpd2 mutants to P. cucumerina. ANOVA detected significant differences at the P<0.05 level.

Figure 5. Histone H3 modifications.

Comparative level of histone modifications of PR-1, PDF1.2a and Actin2 gene promoters as present in leaf samples from Col-0, nrpd1, nrpe1 and nrpd2 plants. (A) Histone H3 Lys4 trimethylation (H3K4me3) on the indicated gene promoters. (B) Histone H3 K9 acetylation (H3K9ac) on the indicated gene promoters. Data are standardized for Col-0 histone modification levels. (C–D) H3K4me3 (C) and H3K9ac (D) modifications on PR-1 and PDF1.2a gene promoters in Col-0 and nrpd2 plants 48 h after inoculation with P. cucumerina. (D) (−) mock inoculated plants, (+) P. cucumerin inoculated plants. Data are standardized for mock inoculated Col-0 histone modification levels. Data represent the mean ± SD; n = 3 biological replicates.

Figure 6. Comparative immune responses of Pol IV and Pol V defective mutants to inoculation with Pseudomonas syringae DC3000.

Growth rates of PsDC3000 in Col-0, nrpd1, nrpe1, nrpd2, and nrpd1 nrpe1 plants. The PsDC3000 disease susceptible mutants ago4-2/ocp11 and npr1 were included for comparison. Data represent the mean ± SD; n = 3 biological replicates.