Botrytis small RNA Bc-siR37 suppresses plant defense genes by cross-kingdom RNAi

ABSTRACT

Pathogens secrete effector proteins to suppress host immune responses. Recently, we showed that an aggressive plant fungal pathogen Botrytis cinerea can also deliver small RNA effectors into host cells to suppress host immunity. B. cinerea sRNAs (Bc-sRNAs) translocate into host plants and hijack the plant RNAi machinery to induce cross-kingdom RNAi of host immune responsive genes. Here, we functionally characterized another Bc-sRNA effector Bc-siR37 that is predicted to target at least 15 Arabidopsis genes, including WRKY transcription factors, receptor-like kinases, and cell wall-modifying enzymes. Upon B. cinerea infection, the expression level of Bc-siR37 was induced, and at least eight predicted Arabidopsis target genes were downregulated. These target genes were also suppressed in the transgenic Arabidopsis plants overexpressing Bc-siR37, which exhibited enhanced disease susceptibility to B. cinerea. Furthermore, the knockout mutants of the Bc-siR37 targets, At-WRKY7, At-PMR6, and At-FEI2, also exhibited enhanced disease susceptibility to B. cinerea, giving further support that these genes indeed play a positive role in plant defense against B. cinerea. Our study demonstrates that analysis of pathogen sRNA effectors can be a useful tool to help identify host immunity genes against the corresponding pathogen.

KEYWORDS:

• Cross-kingdom RNAi
• host-pathogen interactions
• pathogen small RNA effector

Introduction

In plants, pathogen-associated molecular patterns (PAMPs) are recognized by plant microbial pattern recognition receptors (PRRs), leading to PAMP-triggered immunity (PTI). Most PRRs in the model plant Arabidopsis are involved in plant immunity.^{13,4,10,29,40,23} In the plant-pathogen arms race, successful pathogens secrete dozens of effector proteins to modulate host physiology and to undermine PTI, the pathways of which has been well characterized for bacterial and oomycete pathogens.^1,16,17 Bacterial pathogens evolved molecular secretion systems for effector delivery, such as the type-III and type-IV secretion systems,³¹ while the delivery and entry of oomycete and fungal effectors into host cells was just started to be understood. As the second layer of the immune response, plant species have evolved many resistance (R) genes. Most R-genes encode nuclear-binding leucine-rich repeat (NB-LRR)-type receptor-like proteins, which recognize pathogen effectors and induce a robust immune response, named effector-triggered immunity (ETI).^8,16,5

Small RNAs (sRNAs), such as small interfering RNAs (siRNAs) and microRNAs (miRNAs) that trigger RNA interference (RNAi) are important in regulation and fine-tuning of plant immunity genes.^26,19,35 Several plant endogenous siRNAs and miRNAs contribute to the regulation of PTI, such as Arabidopsis miR393, miR160, and miR167;^22,37 ETI, such as natural antisense (nat)-siRNAATGB2,²⁰ Arabidopsis long-siRNA-1,¹⁸ and miRNA863–3p²³; or both PTI and ETI, as in the case of miR393*.³⁸

Even though the regulatory role of endogenous plant sRNAs in plant immunity has been well characterized, little is known about the role of pathogen-derived sRNAs in microbial pathogenicity and host adaptation.³³ Remarkably, the fungal pathogen B. cinerea transmits Bc-sRNA effectors into the host cells to silence important host immunity genes by hijacking the host RNAi machinery.^2,30,36 Bc-sRNA effectors have been found to target important immune-related genes in two host plant species, Arabidopsis and Solanum lycopersicum (tomato). For example, Bc-siR3.2 targets two related genes, Arabidopsis Mitogen-activated protein kinase 1 (At-MPK1) and At-MPK2, as well as the tomato Sl-MPKKK4. MPKs are important proteins involved in signal transduction in the plant immune responses.³⁶ Most Bc-sRNAs that exhibit predicted host plant target genes are derived from long-terminal repeat (LTR) retrotransposon loci in the B. cinerea genome. LTR retrotransposons have been associated with fast-evolving genome regions that related to microbial virulence.^36,35,33 This may result in fast evolutionary turnover of Bc-sRNA effectors which would impact the molecular arms race during host and pathogen coevolution. Our recent study suggested that the fungal plant pathogen Verticillium dahliae, the causal agent of Verticillium wilt, also uses pathogen sRNAs (Vd-sRNAs) as effectors, as Vd-sRNAs were found to be loaded into Arabidopsis Argonaute1 (AGO1) during infection.³⁴ The majority of the predicted host targets of these Vd-sRNA effectors are putative factors of plant immunity. In addition, animal parasites, such as the gastrointestinal nematode Heligmosomoides polygyrus, also secrete sRNAs via extracellular vesicles. These parasite sRNAs can be taken up by mice epithelial cells and suppress host genes that regulate inflammation and immune responses.⁷ Thus, these observations of pathogen- or parasite-induced host gene silencing indicate that such cross-kingdom RNAi - based virulence strategy has evolved in diverse pathogens and pests of both plant and animal systems.

Here, we characterized another Bc-sRNA effector, Bc-siR37, which is predicted to target at least 15 host plant genes in Arabidopsis under stringent prediction criteria.⁶ We confirmed that several of these Arabidopsis target genes were silenced by Bc-siR37. We further verified that three target genes participate in plant defense against B. cinerea: At-WRKY7, At-PMR6 and At-FEI2, which encode an immune-related transcription factor, a pectin-lyase, and a leucine-rich repeat (LRR) receptor kinase, respectively. Thus, identification and characterization of pathogen sRNA effectors is an excellent approach to identify host immunity genes that are involved in plant defense against eukaryotic pathogens.

Results

B. cinerea sRNA Bc-siR37 is predicted to target a large number of host plant genes

Next generation sequencing (NGS) profiling of sRNAs from B. cinerea-infected Arabidopsis leaves, tomato leaves, and tomato fruits led to the identification of Bc-sRNAs, some of which were confirmed to be transported into host cells and silence host immunity genes during the infection process.³⁶ From this data set, we identified another Bc-sRNA effector, Bc-siR37, which was mapped to the Bcin12G_06020 open read frame (ORF) that putatively encodes an ABC-type plasma membrane Bc-ATPase (Fig. 1). Bc-siR37 was identified as a sRNA effector in our previous study based on its multiple predicted host target genes. It was not included in table S3 of³⁶ because we only listed predicted Bc-sRNA effectors with 100 or more reads per million (RPM) of total Bc-sRNAs. Here, we chose to further study Bc-siR37 because it has 15 predicted plant targets when using the same stringent prediction criteria as in³⁶ (Table 1), and several of these targets have already been shown to play a role in host immune responses. Moreover, Bc-siR37 was more enriched in the B. cinerea-infected Arabidopsis libraries as well as the conidiospore fraction library, compared with other sRNAs derived from the Bcin12G_06020 ORF locus (Fig. 1), suggesting that Bc-siR37 is likely functional during host infection.

Figure 1. Bc-siR37 accumulates in B. cinerea-infected Arabidopsis. Bc-sRNA expression profiles at the locus of the Bc-ATPase gene Bcin12G_06020 in B. cinerea mycelium, conidiospore samples, or in the B. cinerea-infected Arabidopsis at 0, 24, 48, and 72 hpi. Bc-siR37 was enriched in infected Arabidopsis at 24 hpi. sRNA reads were normalized to reads per million (RPM) in mycelium and condiospore samples and reads per 10 million (RP10M) in B. cinerea-infected plant samples. The size range of small RNAs that mapped to the Bc-ATPase (Bcin12G_06020) gene locus was 20–35 nt with predominately 21–22-nt, and 24–26-nt classes.

Table 1. Bc-siR37 host targets in Arabidopsis.

Score	Gene ID	Encoded protein
2	AT4G24240.1	WRKY7
2	AT1G68330.1	Protein Kinase, PK
2	AT3G46880.1	Unknown
2.25	AT1G02813.1	DUF538
3	AT2G35620.1	LRR kinase, FEI2
3	AT1G64220.1	TOM7–2 subunit
3	AT2G18350.1	HB24 homeo-box
3	AT5G35980.1	YAK1-like kinase
3	AT3G18050.1	Unknown
3	AT1G63057.2	Unknown
4	AT3G54920.1	Pectin lyase, PMR6
4	AT3G51830.1	ATG5 transmembrane protein
4	AT2G44410.1	RING/U-box
4	AT1G69310.1	WRKY57
4	AT2G36460.2	FBA6

Arabidopsis genes targeted by Bc-siR37 have lower transcript levels upon B. cinerea infection

We first aimed to confirm the accumulation of Bc-siR37 in B. cinerea-infected host samples. Infected Arabidopsis leaf materials were collected at 0, 24, 48, and 72 hours post infection (hpi) for RNA extraction and adaptor ligation-based PCR to amplify Bc-siR37. Bc-siR37 was clearly detected at 0, 24, 48, and 72 hpi and enriched at 24 hpi (Fig. 2A). This observation was consistent with the NGS deep sequencing data (Fig. 1). As a positive control for B. cinerea sRNA production during infection, we detected Bc-siR3.1, a previously characterized Bc-sRNA effector (Fig. 2A).³⁶ Next, if Bc-siR37 has a suppressive effect on predicted host target genes during infection, a lower transcriptional level of these targets would be expected upon B. cinerea infection. Therefore, we performed real-time RT-PCR to quantify transcript levels of At-WRKY7, At-FEI2, At-PMR6, At-ATG5, At-PK, At-TOM7–2, At-DUF538, and At-RING/U-box at 0, 24, 48, and 72 hpi, according to the time points that we had detected the accumulation of the Bc-siR37 by PCR (Fig. 2A) and by NGS data analysis of sRNAs (Fig. 1). Indeed, transcript levels of predicted target genes evaluated here were reduced at 24, 48, and 72 hpi compared with 0 hpi (Fig. 2B). At-PDF1.2 and At-BIK1 were chosen as Botrytis-induced marker genes, which indeed showed strong induction during infection (Fig. 2B).^21,32 Therefore, we assumed that the reduced transcript levels of predicted Bc-siR37 host targets were due to gene silencing rather than plant cell death caused by the pathogen. Nevertheless, comprehensive transcriptional re-programing is common during stress response in plants, which could explain the reduced transcription levels of the predicted target genes during infection. Therefore, we performed transient co-expression assays with Bc-siR37 and two of its target genes, At-WRKY7 and At-FEI2, to confirm that Bc-siR37 can indeed cause gene silencing of the host target genes.

Figure 2. Arabidopsis target genes of Bc-siR37 are suppressed upon infection. (A) Total RNA from B. cinerea and B. cinerea-infected Arabidopsis collected 0, 24, 48, and 72 hpi was isolated for sRNA library construction. Expression of Bc-siR37 was determined by liagation-based sRNA RT-PCR. Bc-siR3.1 was used as a positive control.³⁶ Expression levels of At-ACTIN and Bc-ACTIN were measured as further control for RNA integrity and B. cinerea infection. Both Bc-siR37 and Bc-siR3.1 are 21-nt long, while the sizes of PCR products are 74- and 75-nt, respectively. (B) Relative mRNA levels of the Bc-siR37 host target genes, At-WRKY7, At-FEI2, At-PMR6, At-ATG5, At-PK, At-TOM7–2, At-DUF538, and At-RING/U-box, as well as B. cinerea-induced At-BIK1 and At-PDF1.2 marker genes, were examined by real-time RT-PCR. The error bars indicate the standard deviation (s.d.) of 3 technical replicates. Similar results were obtained in three independent biological replicates.

Bc-siR37 efficiently silences Arabidopsis At-WRKY7 and At-FEI2 genes in transient expression assay

Bc-siR37 targets At-WRKY7 and At-FEI2 at their 5′ UTR region with a nearly perfect complementary sequence alignment (Fig. 3A and 3B). To demonstrate that gene silencing was target site-specific, we conducted an A. tumefaciens-mediated transient co-expression assay of Bc-siR37 with its host target genes using N. benthamiana leaves. We cloned the Bc-siR37 sequence into a plant artificial microRNA vector (amiR-Bc-siR37),²⁸ and the At-WRKY7 or At-FEI2 cDNA sequences into the pEarlygate101 (pE101) binary expression vector that carried a hemagglutinin (HA) and a YFP tag at the C-terminus. We also cloned the synonymous mutations at the sRNA target sites of At-WRKY7 (At-WRKY7m) and At-FEI2 (At-FEI2m) that should be no longer targeted by Bc-siR37 (Fig. 3A and 3B). Silencing efficiencies of Bc-siR37 on At-WRKY7 and At-FEI2 were analyzed at 48 hours post Agrobacterium-infiltration by Western blot using anti-HA antibody. At-WRKY7 and At-FEI2 were highly expressed without Bc-siR37 co-expression (Fig. 3A and 3B). In contrast, when co-expressed with Bc-siR37, there was a clear reduction of At-WRKY7 and At-FEI2 expression (Fig. 3A and 3B), demonstrating sequence-specific gene silencing. Moreover, when we co-expressed the Arabidopsis miRNA At-miR395 instead of Bc-siR37, which had neither sequence homology to At-WRKY7 nor to At-FEI2, no target gene suppression was observed (Fig. 3A and 3B). Finally, when we co-expressed Bc-siR37 with At-WRKY7m or At-FEI2m, the silencing effect was completely abolished (Fig. 3A and 3B).

Figure 3. Bc-siR37 silences the host target genes as revealed by co-expression assays in N. benthamiana. Co-expression of Bc-siR37 and its host targets (HA and YFP-tagged) At-WRKY7 (A) and At-FEI2 (B) in N. benthamiana revealed suppression of the host targets by Bc-siR37. The top panel showes the protein levels of At-WRKY7 and its target site-mutated version At-WRKY7m (A), as well as At-FEI2 and At-FEI2m (B), by Western blot analysis. The bottom panel shows the sequence alignment of Bc-siR37 with its target sites in At-WRKY7/At-WRKY7m (A), and At-FEI2/At-FEI2m (B). (C) Expression of At-WRKY7 and At-WRKY7m after infection with B. cinerea was quantified by confocal microscopy. Error bars indicate the s.d. of 5 pictures, and asterisk indicates the statistical significance of P<0.01 (One-way Anova). Similar results were obtained in 2 independent biological replicates (A-C).

To verify the sequence specificity of YFP-tagged At-WRKY7 silencing during B. cinerea infection of N. benthamiana leaves, we used confocal microscopy coupled with quantitative image analysis. We infected N. benthamiana leaves transiently expressing YFP-At-WRKY7 or YFP-At-WRKY7m with B. cinerea and found that only At-WRKY7 was suppressed at fungal infection sites but not At-WRKY7m (Fig. 3C). Based on these results, we believe that Bc-siR37 indeed silences Arabidopsis host genes At-WRKY7 and At-FEI2.

Transgenic Arabidopsis plants expressing Bc-siR37 exhibited attenuated plant defense against B. cinerea

To further confirm that Bc-siR37 indeed silences host target genes in Arabidopsis, we stably transformed the amiR-Bc-siR37 expression vector into Arabidopsis Col-0 wild type (WT) plants. We obtained in total 13 transformed lines that expressed Bc-siR37 at various levels (Bc-siR37^ox), as shown by Northern blot analysis (Fig. 4A). None of these transformed lines exhibited any obvious morphological or developmental defects. We collected F2 progenies from lines with high expression level of Bc-siR37. To study the silencing effect of constitutively expressed Bc-siR37 on predicted Arabidopsis target genes, we measured transcriptional levels of At-WRKY7, At-FEI2, At-PMR6, At-ATG5, At-PK, At-TOM7–2, At-DUF538, and At-RING/U-box in the Bc-siR37^ox transgenic lines in comparison to WT Arabidopsis plants using real-time RT-PCR. In agreement with previous results that showed lower transcript levels of these genes upon B. cinerea inoculation (Fig. 2B), all tested Arabidopsis target genes showed lower transcript levels in uninfected Bc-siR37^ox lines (Fig. 4B). We thus confirmed that Bc-siR37 alone is capable of suppressing these host target genes in Arabidopsis. Association of Bc-siR37 and plant AGO1 protein was also examined after infection by AGO1 co-immunoprecipitation, and Bc-siR37 was clearly detected in the AGO1-pull down fraction, supporting that Bc-siR37 was loaded into Arabidopsis AGO1 (Fig. S1).

Figure 4. Bc-siR37-expressing Arabidopsis exhibits reduced expression of host target genes and enhanced susceptibility toward B. cinerea. (A) The expression levels of Bc-siR37 in the transgenic Arabidopsis Bc-siR37^ox lines were shown by Northern blot analysis. (B) The mRNA levels of the predicted Bc-siR37 host target genes At-WRKY7, At-FEI2, At-PMR6, At-ATG5, At-PK, At-TOM7–2, At-DUF538, and At-RING/U-box were reduced in Bc-siR37^ox plants. The error bars indicate the s.d. of 3 technical replicates. (C) Arabidopsis Bc-siR37^ox plants were more susceptible to B. cinerea than WT Arabidopsis plants. The lesion size was measured 4 d post infection (dpi) using ImageJ software. Error bars indicate s.d. of 10 leaves. The relative biomass was quantified by real-time PCR, and error bars indicate the s.d. of 3 technical replicates. Asterisks indicate statistical significance (t-test, P<0.01). Similar results were obtained in 3 biological replicates for (B)-(C).

To examine if Bc-siR37^ox transgenic plants are impaired in plant defense against B. cinerea infection, we inoculated Bc-siR37^ox plants and WT plants with a B. cinerea spore suspension. Bc-siR37^ox lines were more susceptible to B. cinerea compared with WT plants (Fig. 4C). Consistent with our previous result,³⁶ the virulence of B. cinerea dcl1 dcl2 was partially restored when infected on Bc-siR37^ox plants (Fig. S2). These results support that host Bc-siR37 target genes have functional roles in plant immune responses against B. cinerea. This notion was further supported by the fact that T-DNA insertion lines of At-WRKY7, At-FEI2, and At-PRM6 showed enhanced disease susceptibility toward B. cinerea infection (Fig. 5A and Fig. S3A). Moreover, the Arabidopsis transgenic At-WRKY7ox lines, which overexpress At-WRKY7, showed enhanced disease resistance against B. cinerea when compared with WT plants (Fig. 5B and Fig. S3B).

Figure 5. The host target genes of Bc-siR37 are involved in plant immunity against B. cinerea. Arabidopsis at-wrky7, at-fei2, and at-pmr6 mutants were more susceptible to B. cinerea (A), while Arabidopsis At-WRKY7^ox lines were more resistant to B. cinerea (B). The lesion sizes were calculated 4 dpi using ImageJ, and error bars represent the s.d. of 10 leaves. The relative biomass was quantified by real-time PCR, and error bars indicate the s.d. of 3 technical replicates. Asterisks indicate statistical significance (One-way Anova for (A), and student's t-test for (B), P<0.01). Similar results were obtained in 3 independent biological replicates.

Discussion

Our previous study only listed 73 Bc-sRNA effectors with at least 100 RPM of total Bc-sRNAs, but there are more Bc-sRNA effector candidates with lower reads in the data set. Most of those potential Bc-sRNA sequences were mapped to retrotransposon regions in the B. cinerea genome.³⁶ In this paper, we characterized a new Bc-sRNA effector, Bc-siR37, which is transcribed from the sense strand of a gene, Bcin12G_06020. Bc-siR37 caught our attention because it was predicted to target many host plant genes, some of which were known to be involved in plant immunity. One of the Arabidopsis target genes, At-ATG5, has already been demonstrated to participate in plant defense against B. cinerea.²⁴ In addition, we confirmed that 3 additional Arabidopsis genes, At-WRKY7, At-FEI2, and At-PMR6, are involved in the plant defense responses against B. cinerea.

Unlike most predicted Bc-sRNA effectors, which are derived from retrotransposon regions, Bc-siR37 is derived from a gene encoding a Bc-ATPase, Bcin12G_06020. Interestingly, two more B. cinerea ATPases, Bc-CCC2 and Bc-PMR1, were proven to be involved in pathogenicity.^27,25 It needs to be clarified whether the Bc-ATPase from Bcin12G_06020 itself is also required for virulence. At least, sRNA Bc-siR37 from this locus is a bona fide virulence factor.

At-WRKY7 is a negative regulator of salicylic acid (SA)-dependent PR1 gene expression in plant defense response to the bacterial pathogen Pseudomonas syringae.²¹ This suggests that At-WRKY7 might positively regulate jasmonate (JA)-dependent defense response due to the antagonism of SA and JA hormone signaling pathways.¹² Indeed, our results indicated a positive role for At-WRKY7 in defense against the necrotrophic pathogen B. cinerea. However, how At-WRKY7 activates JA-responsive defense genes, either directly as a transcriptional activator, or indirectly through inhibition of SA signaling, is still not understood. Yet, another WRKY transcription factor, At-WRKY33, is a positive regulator of plant immunity against B. cinerea, and it has been shown to indirectly regulate the JA pathway signaling.^39,3 Interestingly, another WRKY transcription factor At-WRKY57 is also a target of Bc-siR37. At-WRKY57 was first identified as a transcriptional mediator in response to plant drought tolerance.¹⁵ Moreover, At-WRKY57 participates in activation of JA-responsive genes and actively binds to JASMONATE ZIM-DOMAIN4/8 (JAZ4/8),¹⁴ which suggests that At-WRKY57 is involved in JA-dependent activation of plant defense against B. cinerea. Taken together, Bc-siR37 targets multiple members of the WRKY transcription factor family in Arabidopsis. Similarly, we had found earlier that Bc-siR3.2 targeted the plant MPK family member At-MPK1 and At-MPK2 in Arabidopsis.³⁶ Therefore, we believe that distinct Bc-sRNA effectors target host plant genes that are key factors of the innate immune responses. These Bc-sRNA effectors seem to target either a single pathway through silencing homologous genes to eliminate functional redundancy or by silencing different target genes for the sake of strengthening suppression efficiency.

Many predicted host targets of Bc-sRNA effectors were demonstrated to be involved in plant immune responses against B. cinerea, including At-MPK1, At-MPK2, At-WAK, At-PRXIIF, At-ATG5, At-WRKY7, At-FEI2, At-PMR6, and Sl-MPKKK4.³⁶ Thus, we propose that predicting target genes of pathogen sRNA effectors is a suitable way to identify new host immunity factors. Here, we found that Bc-siR37 has several host target genes with unknown functions (Table 1) but are very likely to be involved in host immunity against B. cinerea. Currently, more and more sRNA data sets from diverse pathogen species are becoming available in public databases. Data mining of pathogen sRNA information provides a rich resource to identify novel immunity gene candidates in many host-pathogen systems, providing a short-cut to laborious forward genetic screens.

Materials and methods

Plasmid construction

The amiR-Bc-siR37 construct was designed and generated according to the website WMD3 (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi). The plasmid RS300 was used as template to amplify three fragments with primer pairs Oligo A and Bc-siR37 I miR (a), Bc-siR37 II miR-a and Bc-siR37 III miR*s (b), Bc-siR37 IV miR*a and Oligo B (c). The a, b, c fragments were used as templates, and together with primers Oligo A and Oligo B to amplify fragment d, which exchanged Bc-siR37 into At-miR319a backbone. Fragment d was cloned into pENTR/D-TOPO (Invitrogen, Carlsbad, CA), and next into the destination vector pEG100 by LR reactions (Invitrogen, Carlsbad, CA). Primer sequences are listed in supplementary materials.

At-WRKY7, At-WRKY7m, At-FEI2, and At-FEI2m overexpression vectors: Arabidopsis cDNAs were used as template with primers At-WRKY7-FOR and At-WRKY7-REV, At-WRKY7m-FOR and At-WRKY7-REV, At-FEI2-FOR and At-FEI2-REV, At-FEI2m-FOR and At-FEI2-REV to amplify the cDNA sequences of At-WRKY7, At-WRKY7m, At-FEI2, and At-FEI2m, respectively. At-WRKY7, At-WRKY7m, At-FEI2, and At-FEI2m cDNA fragments were cloned into pENTR/D-TOPO (Invitrogen, Carlsbad, CA), and finally into the destination vector pEG101 by LR reactions (Invitrogen, Carlsbad, CA). All primer sequences are listed in supplementary materials.

Plant materials

Arabidopsis thaliana ecotype Columbia (Col-0) and N. benthamiana were grown at 22°C under 12 h light per day. All mutants used in this study were in the Col-0 background. The at-wrky7 mutant line and At-WRKY7ox line were provided by Dr. Zhixiang Chen's laboratory, and at-pmr6 (CS6354) and at-fei2 (SALK_083958) T-DNA insertion lines were purchased from the Arabidopsis Biological Resource Center (Ohio State Universide, USA).

To generate the Bc-siR37^ox lines, the amiR-Bc-siR37 construct was transformed into A. tumefaciens GV3101, which was then used to transform Arabidopsis by the flower dipping method as described previously.⁹

Adaptor ligation-based PCR method to amplify sRNAs

Total RNA from different tissues was extracted using TRIzol (Invitrogen, Carlsbad, CA). RNA extracts were size-separated on a 14% denatured (8 M Urea) polyacrylamide gel, and the size fractions of 18–30 nt in length were cut out of the gel and purified. The purified sRNA was ligated to 3′ RNA adaptors using truncated T4 RNA ligase 2 (NEB, Ipswich, MA). The ligation products were separated on a 14% denatured polyacrylamide gel and RNA size fractions from 30–50 nt were cut out of the gel and purified. The 3′ adaptor ligated RNA products were ligated with a 5′ RNA adaptor by T4 RNA ligase 1 (NEB, Ipswich, MA). Upon gel size fractionation and purification of the 60–80 nt, the final RNA ligation product was converted into first strand cDNA. The cDNA was amplified using adaptor specific primers SBS5′ and SBS3′. The 116 nt amplification product was purified and used as template for Bc-siR37 amplification using the Bc-siR37 lib forward primer and 3′ adaptor reverse primer SBS3'. All primer sequences are listed in supplementary materials.

Measure mRNA levels of target genes by real-time RT-PCR

Total RNA was extracted from plant tissues using TRIzol (Invitrogen, Carlsbad, CA), and treated with DNase I enzyme (Roche, Basel, Switzerland). The purified DNase I-treated RNA was converted into first strand cDNA using SuperScript III (Invitrogen, Carlsbad, CA). The cDNA was used as template for real-time RT-PCR using the Bio-Rad IQ5 thermocycler (Bio-rad) and the iQ SYBR Green Supermix (Bio-Rad, Hercules, CA), and gene-specific RT-PCR primers (Table S1). The PCR program performed as followed: 95°C at 2 min and 45 cycles of 95°C at 30 s, 55°C at 30 s, 72°C at 30 s, and fluorescence was measured at the end of extension step for 30 s.

Co-expression assay

A. tumefaciens carrying amiR-Bc-siR37 or amiR395 was co-infiltrated with At-WRKY7, At-WRKY7m, At-FEI2, or At-FEI2m into leaves of four-week-old N. benthamiana plants. Two days after infiltration, leaf samples were collected for Western blot analysis or confocal microscopy. Quantitative microscopy analysis was performed as described previously.³⁶

B. cinerea infection assay

B. cinerea spores were collected from 10 d old B. cinerea cultured on malt extract agar plates and diluted to 2 × 10⁵ spores per milliliter in inoculation medium (10 mM sucrose, 10 mM KH₂PO₄, Tween-20 0.025%) for infection. A 15 µl spore solution drop was placed onto the center of four-week-old Arabidopsis leaves. Photos were taken with a digital camera and analyzed using ImageJ software to measure the lesion size. The calculation of relative DNA contents of B. cinerea to Arabidopsis (or relative biomass) was described previously.¹¹ For infection of N. benthamiana leaves, B. cinerea spores were drop inoculated at the abaxial (bottom) side of Agrobacterium-infiltrated leaves.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Dr. Zhixiang Chen for at-wrky7 mutant and At-WRKY7ox line seeds. We thank Yifan Lii for editing the paper.

Funding

This work was supported by grants from the National Institute of Health (R01 GM093008) and National Science Foundation (IOS-1257576, IOS-1557812) awarded to HJ.