miR403a and SA Are Involved in NbAGO2 Mediated Antiviral Defenses Against TMV Infection in Nicotiana benthamiana

RNAi (RNA interference) is an important defense response against virus infection in plants. The core machinery of the RNAi pathway in plants include DCL (Dicer Like), AGO (Argonaute) and RdRp (RNA dependent RNA polymerase). Although involvement of these RNAi components in virus infection responses was demonstrated in Arabidopsis thaliana, their contribution to antiviral immunity in Nicotiana benthamiana, a model plant for plant-pathogen interaction studies, is not well understood. In this study, we investigated the role of N. benthamiana NbAGO2 gene against TMV (Tomato mosaic virus) infection. Silencing of NbAGO2 by transient expression of an hpRNA construct recovered GFP (Green fluorescent protein) expression in GFP-silenced plant, demonstrating that NbAGO2 participated in RNAi process in N. benthamiana. Expression of NbAGO2 was transcriptionally induced by both MeSA (Methylsalicylate acid) treatment and TMV infection. Down-regulation of NbAGO2 gene by amiR-NbAGO2 transient expression compromised plant resistance against TMV infection. Inhibition of endogenous miR403a, a predicted regulatory microRNA of NbAGO2, reduced TMV infection. Our study provides evidence for the antiviral role of NbAGO2 against a Tobamovirus family virus TMV in N. benthamiana, and SA (Salicylic acid) mediates this by induction of NbAGO2 expression upon TMV infection. Our data also highlighted that miR403a was involved in TMV defense by regulation of target NbAGO2 gene in N. Benthamiana.


Introduction
Virus diseases threaten crop production and agriculture globally. Plant evolved and developed sophisticated mechanisms to defend against virus infections. RNA interference (RNAi) is an antiviral defense mechanism in plants. The history of our understanding of RNAi for viral defense can be traced back to 1928, when Wingard reported the recovery of virus infected plant from later virus infection [1]. RNAi phenomenon was then reported in petunia in 1990 and was termed as "co-suppression" and later as post transcriptional gene silencing (PTGS) [2]. In 1993, Lindbo et al. suggested a working mechanism for RNAi while studying gene silencing phenomena in transgenic plants [3]. In 1998, Fire et al. silenced endogenous gene expression in Caenorhabditis elegans by injection of double-stranded RNA (dsRNA), which was therefore identified as the molecule mediating sequence specific inhibitory effects during RNAi process [4]. In 1999, Hamilton and Baulcombe found that 25-nucleotide antisense RNA served as the intermediate for RNAi [5]. RNAi was also found in fungus and virus infected plant, and was coined as quelling and virus-induced gene silencing (VIGS) respectively [6][7][8]. Briefly, and 70% relative humidity. Three-leaf stage seedlings were used for following experiments. 0.5 mM MeSA (Methylsalicylic acid) in water solution was sprayed on the leaves and the leaves were collected one day after treatment. Water solution was sprayed as mock treatment. For GFP visualization, TMV-GFP infected plants were illuminated with a hand-held UV lamp at 320 nm or under Nikon SMZ18 microscope (Nikon, Tokyo, Japan).

SA Content Measurement
The content of endogenous salicylic acid (SA) was determined by UV Vis Spectrophotometry. A set of standard solutions, containing 5, 10, 20, 30, and 40 µg/mL of SA, were prepared in 60% ethanol solution and one of the standard solution was used for measuring maximum absorption wavelength within a 280-320 nm wavelength range (5 nm interval), which was used as a subsequent detection wavelength. The maximum absorption wavelength was 310 nm and the linear regression equation according to the absorbance of five standard solutions was deduced. N. benthamiana leaves infected with TMV and mock solution were collected and frozen in liquid nitrogen after five hours. The collected leaves were ground in liquid nitrogen and soaked with 5 mL 60% ethanol, and the supernatant after centrifugation was used to detect absorbance at 310 nm, which was then converted to SA concentration according to the standard curve [38].

Construction of hpGFP Vector
For silencing of GFP gene, GFP interfering sequences were amplified by Forward primer 5 -TGCGGGATATCGGACGACGGGAACTACAAGA-3 and reverse primer 5 -TGCGGGATATCAAAGGGCAGATTGTGTGGAC-3 from genomic DNA of GFP16c line. The fragment was cloned into pQBV3 entry vector by EcoRV and then was subcloned into binary vector pCB2004B to produce hairpin construct for insert GFP fragment.

Construction of amiR Based Silencing Vector for NbAGO2 Gene
For construction of viral pCV (CaLCuV)-based (cabbage leaf-curl virus) amiR vector, following oligonucleotides were designed by amiR design function in www.benthgenome.com website. amiR1: The above primers were used to amplify amiR fragment using Arabidopsis thaliana miR159b as template backbone and the amplicons were cloned into pCVA vector by XbaI/KpnI sites. The same amplicons were cloned into pBI121 vector by XbaI/KpnI sites to produce non-viral binary vector based amiR constructs. from cDNA of N. benthamiana and the amplicon was cloned into pBI121 vector by BamHI/SacI sites.

Construction of STTM-Based Inhibitory Vectors for Endogenous microRNAs
For development of STTM (Short tandem target mimic) constructs for miR390a, miR393a and miR403a, the following sequences and their complementary sequences were designed and synthesized. miR390a, 5 -ctcgagGGTGCTATCCCCTATCCTGAGCTTGTTGTTGTTGTTATGGTCTAATTTAAATAT GGTCTAA AGAAGAAGAATGGTGCTATCCCCTATCCTGAGCTTgaattc-3 ; miR393a, 5 -ctcgagCCAA AGGGCTAATAGCATGATGTTGTTGTTGTTATGGTCTAATTTAAATATGGTCTAAAGAAGAAGAAT CCAAAGGGCTAATAGCATGATgaattc-3 ; and miR403, 5 -ctcgagAGTTTGTGCCTAGTGAATCTAA GTTGTTGTTGTTATGGTCTAATTTAAATATGGTCTAAAGAAGAAGAATAGTTTGTGCCTAGTGAA TCTAAgaattc-3 . The restriction endonuclease sites are shown in lower-case letter, and the STTM spacer sequences were underlined. The oligonucleotides and complementary sequences were annealed to produce dsDNA and this dsDNA was then ligated into the pGreen-GUS-Competitor vector by XhoI and EcoRI respectively.

Plant Inoculation, Gene Silencing and Virus Infection
The above constructs were transformed into Agrobacterium tumefaciens GV3101 competent cells. The transformants were grown to optical density OD 600 = 1 at 28 • C, cells were collected by centrifugation and resuspended in inoculation buffer (10 mM 2-(N-Morpholino)ethanesulfonic acid (MES), 10 mM MgCl 2 , 200 µM acetosyringone). OD 600 values of agrobacterium inoculation solution were all adjusted to 1.0 unless stated otherwise. The inoculation solution was leaf-infiltrated into three-leaf stage N. benthamiana by needleless syringe. For pCV-based silencing experiment, pCVB was simultaneously transformed and co-infiltrated with recombinant pCVA vectors. For virus infection, infectious clone of TMV expressing green fluorescent protein (TMV-GFP), pJL24, was infected as the same way. For pJL24, OD 600 value of Agrobacterium inoculation solution was 0.1. In all experiments, inoculation buffer with empty vector was used as mock treatment.

Total RNA Extraction and cDNA Preparation
Leaf tissues were grounded in liquid nitrogen by using mortar and pestle. The powder was homogenized in Trizol reagent (CWBio, Beijing, China) and extracted following manufacturer's protocols. DNase was used to remove contaminated genomic DNA. Total RNA quality was confirmed by measurement of OD 260 /OD 280 using NanoDrop2000 Spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis. One µg of total RNA was used as template and cDNA was synthesized by random hexamer using GoScript Reverse Transcription System (Promega Corporation, Madison, WI, USA).

Assessment of NbAGO2 Expression by qRT-PCR Experiment and Semi-Quantitative RT-PCR
The cDNAs were subjected to PCR experiment. The PCR reaction was run on qTOWER 2.2 (Analytik Jena AG, Jena, Germany) equipment. Fast SYBR Green Master Mix (ThermoFisher Scientific, Waltham, MA, USA) was used for evaluation of NbAGO2 level by qRT-PCR. The primer sets 5 -ATGTGAAATGGTACGGGCTGAA-3 and 5 -CAACAAGGTTCCACTGGCATTT-3 were used for NbAGO2. As an internal control, steady state level of N. benthamiana GAPDH gene was evaluated by primer set 5 -CTGACAAGGACAAGGCTGCT-3 and 5 -AAGCAGCTCTTCCACCTCTC-3 . The ratio of mRNA level between treatment and mock samples was determined by the 2 −∆∆CT method. At least three biological replicates with two technical replications were carried out, and the differences were evaluated by t-test.

Northern Hybridization Experiment
Ten µg of total RNA was resolved by agarose gel electrophoresis. RNAs were transferred from the gel to nylon membrane (Amersham Hybond-N + , positively charged) (GE Healthcare Life Sciences, Pittsburgh, PA, USA) by an electrokinetic capillary apparatus. 32 P or digoxigenin-labelled probes were prepared by the methods described as previously, and hybridization and signal visualization were performed as per the literatures [39,40]. For N. benthamiana NbPR1a gene, a contig sequence with transcript ID of JN247448 from N. benthamiana sequence database (www.benthgenome.com) was used and the primer pairs 5 -GGATGCCCATAACACAGCTC-3 and 5 -CCTAGCACATCCAACACGAA-3 were used for amplification of 313bp probe sequence. For N. benthamiana β-actin gene, a partial cDNA sequence (JQ256516.1) was used and the primer pairs 5 -CCCAAAGGCTAATCGTGAAA-3 and 5 -GCAGCTTCCATTCCAATCAT-3 were used for amplification of 483bp probe sequence. For TMV detection, nucleotide encoding capsid protein was used and the primer pairs 5 -CAAGCTCGAACTGTCGTTCA-3 and 5 -GACCAGAGGTCCAAACCAAA-3 were used for amplification of 352bp probe sequence.

Transient Silencing of NbAGO2 in N. benthamiana by amiR-NbAGO2
Artificial miRNA recently emerged as a newly developed technology for gene silencing. NbAGO2 amiR sequences were designed in www.benthgenome.com website and were used to amplify amiR fragment using Arabidopsis thaliana miR159b as template backbone ( Figure 1A, Supplementary Figure  S1). An overexpression vector for amiR-NbAGO2 was constructed in CaLCuV (cabbage leaf-curl virus, Geminivirus) -based vector pCVA under 35S promoter [37]. The pCVA-amiR-NbAGO2 was transformed into Agrobacterium tumefaciens strain GV3101 and agro-infiltrated into N. benthamiana leaves together with pCVB. We used three different regions of NbAGO2 for amiR targeting, and named them amiR-NbAGO2-1, amiR-NbAGO2-2 and amiR-NbAGO2-3 respectively ( Figure 1B). Three days later, silencing effects on NbAGO2 expression were determined by qRT-PCR experiment. As shown in Figure 1C, overexpression of amiR-NbAGO2 significantly down-regulated the level of NbAGO2 by 53-74%. The most pronounced effect was observed for amiR-NbAGO2-2. The results demonstrated the efficacy of amiR constructs for knockdown of NbAGO2 expression, and different amiR varied as regards to silencing efficacy. Therefore, we used amiR-NbAGO2-2 for following experiments. As inoculation of pCV interfere with subsequent infection of other viruses like TMV-GFP (Supplementary Figure S2), we then used pBI121-based binary vector to overexpress amiR. Time course experiment showed that silencing efficiency of pBI121-amiR was most evident at 6 DPI (days post infection) ( Figure 2). In the following experiments, we used pBI121-amiR-NbAGO2-2 for NbAGO2 silencing experiments. To verify the effectiveness of amiRNAs on NbAGO2 expression, another internal control, GAPDH, was used and the results were similar to the results from NbActin as internal control (Supplementary Figure S3).  pBI121-amiR2 construct was transformed into GV3101, infiltrated on N. benthamiana leaves, and the leaves were collected at different DPI. NbAGO2 expression was determined by quantitative real-time PCR. Plants inoculated with empty pBI121 vector (EV) served as control treatment. NbAGO2 expression in control was arbitrarily designated as 1. NbActin gene served as the internal control. Significant differences between control and treatment groups were calculated by Student's t-test. Double and three asterisks indicated significant difference at p < 0.01 and p < 0.001, respectively.

Silencing of NbAGO2 Compromised dsRNA-Mediated Gene Silencing in N. benthamiana
To confirm that NbAGO2 is functionally involved in RNAi process, a GFP16c N. benthamiana line which overexpresses green fluorescent protein (GFP) was used. First, we used pCB2004B-hpGFP to transiently silence GFP expression in GFP16c plant. The result showed that GFP expression was largely silenced by transient overexpression of hpGFP by agroinfiltration, and both GFP fluorescence and GFP mRNA levels decreased by hpGFP expression. One should note that GFP signal in GFP16c plant was mostly localized in vascular tissue. Next, we used this system to investigate the involvement of NbAGO2 in RNA silencing. Down-regulation of NbAGO2 by amiR-NbAGO2 recovered both the GFP fluorescence and GFP mRNA levels in the hpGFP expressed GFP16c plant, while in the control plant GFP was significantly silenced ( Figure 3A,B). The loss of silencing by NbAGO2 knockdown via amiR-NbAGO2 transient expression verified the participation of NbAGO2 in RNAi process.

Expression of NbAGO2 was Induced by Both MeSA Treatment and TMV Infection
Salicylic acid (SA) is reported to be broadly engaged in virus infection and defense response. We also investigated the involvement of SA in the TMV infection process in N. benthamiana. Exogenous MeSA was sprayed on three-leaf stage N. benthamiana seedlings and the seedlings were collected 2, 10 and 24 h post spray. To verify the effectiveness of SA treatment, expression of SA marker gene NbPR1a was determined by Northern blot experiment. As shown in Figure 4A, NbPR1a was strongly induced by MeSA treatment after 10 h, demonstrating that foliar application of MeSA is effective in N. benthamiana. Quantitative real-time PCR was performed to determine the mRNA level of NbAGO2. Under MeSA treatment, transcript level of NbAGO2 was shortly elevated at 10 h by 2.4-fold and then declined to basal level by 24 h (Figure 4B). We then evaluated the expression of NbAGO2 in response to TMV infection. NbAGO2 was significantly induced by TMV infection during 1-4 DPI ( Figure 5A). In tobacco, SA was responsible for plant resistance to virus infection, especially of TMV. Our data showed that SA level was significantly elevated by TMV infection (Figure 5B), and NbPR1a was also induced by TMV infection in N. benthamiana ( Figure 4A).

Down-Regulation of NbAGO2 Promoted TMV Infection
As shown above, NbAGO2 was responsive to TMV infection and MeSA treatment. Involvement of NbAGO2 in TMV infection was further investigated as follows. NbAGO2 was first silenced by pBI121-amiR-NbAGO2 agroinfiltration, and after one day the same leaf region was agroinfiltrated with TMV-GFP, an infectious TMV engineered to express green fluorescent protein, which was retained in the infiltrated area for at least 6 days. The inoculated leaves were monitored daily for the appearance of GFP fluorescence. As shown in Figure 6A, silencing of NbAGO2 caused stronger accumulation of virus (GFP fluorescence) as compared to control empty vector (pBI121) inoculated leaves. In accordance with this, overexpression of NbAGO2 (Ox-NbAGO2) resulted in weaker virus accumulation, and this effect was abolished by simultaneous amiR-NbAGO2 expression. In addition, virus spread was reduced by treatment with MeSA ( Figure 6B). TMV RNA level was also determined by Northern blot experiment with CP (Capsid protein) region-specific probe, and the result was consistent with GFP fluorescence data ( Figure 6C,D). Taken together, we concluded that NbAGO2 was an antiviral factor effective against TMV, and SA was implicated in this process by inducing NbAGO2 expression, at least to some extent. Northern blot experiment. The same treated samples from above were subjected to detection of TMV RNA. Total RNAs were isolated and TMV RNA level was determined by Northern blot experiment using capsid protein sequence specific probe labelled with 32 P. NbActin served as internal control and the mRNA level of NbActin was evaluated in the same way using 32 P labelled NbActin specific probe (lower panel). The band corresponding to TMV genomic DNA was indicated by the arrowhead. The number below the panel indicates the ratio of the blot band intensity between TMV and NbActin, in which the ratio in control lane was arbitrarily designated as 1.

STTM_miR403a Suppressed TMV Infection
As a core element in RISC formation, AGO2 plays an important role in RNAi-mediated disease resistance response. We therefore hypothesized that some miRNA, especially which participate in RNAi, may have an effect in the process. We chose three miRNAs to further verify this. Among them, miRNA390a and miRNA393a were reported to influence the activity of AGO2 by binding to it [41], while miRNA403a was reported to mitigate AGO2 expression by posttranscriptional regulation [42]. Short tandem target mimic (STTM) is a complementary RNA which blocks small RNA functions in plants by sponge activity. Therefore, we designed an STTM inhibitory vector for each miRNA to examine functions of these miRNAs in virus defense according to the methods described by Tang et al. [43]. After co-inoculation with STTM constructs and TMV-GFP via agroinfiltration in N. benthamiana plants, the GFP fluorescence was observably weakened in STTM_miR390a, STTM_miR393a, and STTM_miR403a co-infected area at 6 DPI compared to pGreen_GUS_competitor, which served as the control construct ( Figure 7A). Subsequently, fluorescence intensity was quantified, and the GFP intensity of STTM co-expressed groups, especially that of STTM_miR403a, showed statistical differences compared with the control (Figure 7B). In the subsequent analysis, we found that miR403a had a target site in 3 UTR of NbAGO2 transcript (Supplementary Figure S4). STTM_miR403a expression reduced NbAGO2 mRNA level ( Figure 7C), while STTM_miR390a, STTM_miR393a expression did not alter NbAGO2 mRNA level. Our data showed the regulatory role of miR403a on the expression of NbAGO2 as well as its role in antiviral response.

Discussion
SA-mediated defenses were involved in plant response against a variety of pathogen types including viruses, bacteria and fungi [44]. Involvement of SA in plant response against virus infection was first established for the well-known interaction between TMV and its resistance gene N from tobacco [45]. Furthermore, SA was reported to be inductively accumulated after virus infection, highlighting the importance of this plant hormone in compatible plant-virus interactions [46,47]. Both RNA viruses and DNA viruses were reported to trigger SA signaling pathway [48,49]. In fact, accumulation of SA after pathogen attack and induction of hallmark PR genes were commonly observed [50,51]. Infection of a diverse range of plant pathogens activated SA signaling pathway and subsequent hypersensitive reaction (HR) or systemic acquired resistance (SAR) [51]. RNAi is a well-known antiviral defense mechanism in plants. Cross-talk of RNA silencing and SA triggered defense responses were manifested by several reports. First, N. benthamiana AGO4 protein was involved in translational repression and was indispensable for SA-mediated defense response by N gene [17]. Second, silencing suppressor proteins encoded by virus genome influence both RNAi and SA-induced plant defenses against viruses [52][53][54]. Third, expression of some RNA silencing component genes are influenced by SA treatment. For example, among plant RNA silencing components, activity or expression level of RdRp1 from a variety of plant species including Arabidopsis thaliana, N. benthamiana, maize, cotton, potato and tobacco were elicited by SA or virus infection, as led to activation of antiviral defense responses [26,27,[55][56][57][58][59]. Our data showed that foliar application of MeSA induced NbAGO2 expression and activated defense against TMV, as is consistent with results from other studies [60,61]. We used TMV infectious clone pJL24 and it was retained in the infiltrated area within our detection time. Induction of NbAGO2 expression by SA indicated that cross-talk of RNA silencing and SA-mediated antiviral defenses were mediated by RdRp or AGO2 in some plant species like N. benthamiana [62]. As for the TMV defense reaction, NbRdRp1 was reported to function in systemic leaves but not on directly-inoculated leaves [27,55,63], while NbAGO2 played TMV defense role in inoculated leaves in our results. Induction of RNAi pathway component genes by corresponding pathogens were reported in several manuscripts. In Arabidopsis thaliana, AGO2 was induced by bacterial pathogen Pseudomonas syringae, virus pathogens TCV, CMV and also by virus silencing suppressor proteins [41,53,64]. RdRp1 expression was induced by TMV infection [55]. Expression of DRB4 was induced by infection of viruses such as tobacco yellow mosaic virus (TYMV, Tymovirus), TCV, CaMV [65] and TCV [66,67]. In rice, RSV infection induced expression of AGO18 and AGO2, while RDV infection induced expression of several OsRDR genes but not OsAGO genes [68]. Here, we showed that expression of NbAGO2 was induced by TMV infection in virological model species N. benthamiana. NbAGO2 was shown to be involved in both RNAi and virus defense responses here. NbAGO2 was reported to mediate defense against tomato bushy stunt virus (TBSV, Tombusvirus) [32]. CRISPR/Cas9 mediated inactivation of NbAGO2 compromise defenses against PVX, TuMV and TCV, but not Tombusvirus and CMV infections [33]. Transgenic approach showed that AGO2 defends against TBSV, TMV and PVX, but not foxtail mosaic virus (FoMV, Potexvirus) [69]. Our data add weight to the present evidence that AGO2 is a broad spectrum antiviral gene whose activity is achieved by RNAi mechanism. Artificial miRNA was shown to be effective to silence variety of genes in a diversity of plant species including N. benthamiana [70], Arabidopsis thaliana [71], N. tabacum [72], Oryza sativa [35], Glycine max [73], Solanum melongena [74], Medicago truncutula [75], Triticum aestivum [76], Vitis vinifera [77], Solanum tuberosum [78] and Solanum lycopersicum [79]. Here, we also showed that artificial miRNA was an effective way to silence N. benthamiana AGO family genes via transient expression from both virus vector and binary vector. As infection of virus vector (pCV) hindered subsequent TMV infection, we used pBI121 for expression of amiRNA and its effects on TMV infection. It is interesting to note that AGO2 itself is targeted by miR403a in plants [80], and here we showed for the first time that attenuation of miR403a inhibition on NbAGO2 by STTM technology influenced TMV infection. As for the miR390a and miR393a, both of them were reported to bind to NbAGO2 protein competitively [41], therefore we speculated that this is the reason that they play similar role with miR403a in our experiments. Paudel et al. reported that NbAGO2 expression was induced by tomato ringspot virus (ToRSV; Secoviridae) isolates ToRSV-Rasp1 or ToRSV-GYV, and silencing of NbAGO2 elevated ToRSV accumulation [81]. This indicated that NbAGO2 was involved in plant responses toward different families of viruses.

Conclusions
The role of NbAGO2 gene against TMV infection was investigated in N. benthamiana. NbAGO2 was demonstrated to be involved in RNAi process in N. benthamiana. NbAGO2 mRNA level was elevated by both TMV infection and MeSA treatment. Suppression of NbAGO2 gene by amiR technology compromised plant resistance against TMV infection. Inhibition of endogenous miR403a, a predicted regulatory microRNA of NbAGO2, reduced TMV infection. Taken all this together, we conclude that SA participates in NbAGO2 mediated TMV response and miR403a is also involved in this response by post-transcriptionally regulating NbAGO2 expression (Figure 8). The sequences for amiR target sites were shown on top by lines with different colors, Figure S2: N. benthamiana leaves were first inoculated with pCV (both pCVA and pCVB), and then were infected with TMV-GFP (pJL24) one day later. As a control, pCambia1301, the backbone of pCV infectious clone used here, was inoculated. The result showed that first inoculation of pCV impeded TMV-GFP infections, Figure S3: The experiments in (A) and (B) were performed the same as Figures 1 and 2, respectively. The only difference is Actin gene was used as internal control here instead of GAPDH in Figures 1 and 2, Figure S4: Sequence alignment between miR403a and target NbAGO2 gene from 17Wt Nicotiana Benthamiana. "|" indicates perfect match, while gap shows mismatches.