Pathogen-Derived Resistance (PDR) is a specific resistance of plants to pathogens by introducing a pathogen into the plant genome. It is widely shown that the PDR to virus infection are relevant to RNA silencing known well for homology-dependent selectively degradation of RNA (Baulcombe, 2005; Hull, 2002). The RNA silencing machinery recognizes several features of viral infections involving the formation of double-stranded (ds) RNA and initiates a response that degrades viral RNA and eventually enables the plant to recover from virus infection. The dsRNA triggers degradation of homologous RNAs and is diced into small interfering (si) RNAs of 21–25 nucleotides (nt) in length. The siRNAs then act as a guide to recognize complementary RNAs for their degradation (Voinnet, 2005; Waterhouse et al., 2001). RNA silencing is also activated by transgenes expressing inverted-repeat (IR) structures that produce dsRNA (Chuang and Meyeriwitz, 2000; Waterhouse et al., 2001) or aberrant transcripts that could be templates for a cellular RNA-dependent RNA polymerase activity (Lipardi et al., 2001). In addition, RNA silencing can be induced by expression of hairpin (hp) RNA in plants, and a variant of this construction which also encodes a spliceosomal intron inserted between the hpRNA arms (so called intron–hpRNA) induced RNA silencing with almost 100% efficiency when directed against RNA virus or endogenous plant genes (Pandolfini et al., 2003; Smith et al., 2000). Transient expression triggered by infiltration of Agrobacterium tumefaciens cultures (so called agroinfiltration) into leaf tissue has been widely used to induce RNA silencing (Johansen and Carrington, 2001; Llave et al., 2000; Voinnet et al., 1999).
Cucumber mosaic virus (CMV) is One of prevalent plant viruses all over the world and has the widest host range of over 885 plant species in 65 families (Palukaitis et al., 1992). CMV, the type species of the genus Cucumovirus in the family Bromoviridae, has a tripartite genome of positive- sense single-stranded RNAs, designated as 1, 2 and 3 in order of decreasing size (Peden and Symons, 1973). RNA2 codes for the 2a protein, which is an RNA-dependent RNA polymerase of replication complex, whereas, RNA1 codes for the 1a protein, another subunit of CMV replicase complex (Hayes and Buck, 1990). RNA3 encodes for two proteins involved in viral movement and encapsidation (Canto et al., 1997; Kaplan et al., 1997). CMV is transmitted by aphids in a non-persistent manner. CMV virion is icosahedral particles, composed of one homologous protein (coat protein; CP). Early classification of CMV strains using serological properties and host responses divided into two subgroups based on different criteria (Palukaitis, et al., 1992), now named subgroup I and subgroup II according to molecular analyses of the genomic RNAs (Owen and Palukaitis, 1988). Subsequently, based on nucleic acid hybridization, restriction fragment length polymorphism and phylogenetic tree analysis, CMV strains have been definitely divided into subgroup I and subgroup II. The members of subgroup I are more prevalent than those of subgroup II in nature (Gallitelli, 2000; Yu et al., 2004). The members of subgroup I have been further divided into subgroup IA and subgroup IB, based on phylogenetic analyses of 5´ non-coding region (NCR) of RNA3 (Roossinck et al., 1999). Isolates from subgroup IA and II have been found all over the world, but with one exception, isolates of subgroup IB are all from East Asia (Palukaitis and García-Arenal, 2003). The members of CMV subgroup IA and subgroup IB are more harmful virus for pepper production in Korea (Cho et al., 2007; Choi et al., 2001; Choi et al., 2005) than those of subgroup II.
RNA silencing has been efficiently used to generate resistance against plant viruses in many ornamental plants (Bucher et al., 2006; Hammond et al., 2006; Tenllado et al., 2004) and in different host systems to obtain resistance against several other viruses (Abhary et al., 2006; Di Nicola-Negri et al., 2005; Lennefors et al., 2006; Pooggin et al., 2003; Tenllado et al., 2003; Vanitharani et al., 2003). Particularly, transgenic expression of pathogen-derived sequences encoding hpRNAs that undergo to an efficient RNA silencing is a novel agricultural control strategy to obtain virus-resistant plants (Smith et al., 2000). However, in non-transgenic plants, it is not known if transient expression of a hpRNA can inhibit multiplication and spread of a widely replicating CMV.
In this study, we show that transient expression of a hpRNA construct using agroinfiltration allows high resistance of Nicotiana benthamiana to CMV. RNA silencing is responsible for this resistance to CMV and this approach makes it possible to construct transgenic crops conferring resistance against CMV.
MATERIALS AND METHODS
Plasmid construction
A cDNA fragment homologous to 162 bp of 3´ NCR portion of RNA3 of CMV strain Fny (CMV-Fny) was synthesized by RT-PCR using the primers showing several conserved sequences sites in 3′ NCRs of RNA1 in all CMV subgroup IA isolates, as described previously (Palukaitis et al., 1992; see also Fig. 1). Briefly, cDNA synthesis using CMV genomic RNAs (approximately 100 ng) purified from virions was performed in a 20 μl volume of 1x Superscript III reaction buffer (Invitrogen, USA), containing 0.5 mM dNTP mix, 5 mM DTT, 40 U RNaseOut, and 200 U of SuperScript III Reverse Transcriptase (Invitrogen, USA) 20 pmole gene-specific reverse primer (attCMVNCR- Reverse; 5’- GGGACCACTTTGTACAAGAAAGCTGGGTGTACACGGACCGAAGTCCTTCC) complementary to positions 2123-2144 (bold) in the RNA3 of CMV-Fny. A Gateway™ sequence (underline) was created in the primer attCMVNCR-Reverse to facilitate further cDNA cloning. The tenth volume of the RT reaction was used for PCR amplification with the primer attCMVNCRReverse and a forward primer (attCMVNCR-Forward; 5’- GGGGACAAGTTTGTACAAAAAAGCAGGCTGGGTCTGAAGTCACTAAACACA) corresponding to positions 1983-2002 (bold) in the RNA3 of CMV-Fny. A Gateway ™ sequence (underline) was created in the primer attCMVNCR-Forward to facilitate further cDNA cloning. PCR amplification was performed in 50 μl of 1x Platinum® Taq DNA polymerase reaction buffer containing 1 mM MgCl2, 0.2 mM dNTP, 10 pmole of each primer and 1 U of Platinum ® Taq DNA polymerase according to manufacturer’s instructions (Invitrogen, USA). The thermal cycles were as follows: 2 min at 94°C (1 cycle); 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C (35 cycles).
In the first cloning step, amplified RT-PCR products were cloned into pDONR207 vector using BP clonase®, according to manufacturer’s instructions (Invitrogen, USA). Then, RT-PCR fragments were cloned into pK7GWIWG2( I) (Karimi et al., 2002) using LR clonase®, generating pIR-CMVNCR. This second ligation positioned the PCR product in inverted orientation with respect to first cloned fragment, yielding IR sequences separated by Arabidopsis intron cloned in the pK7GWIWG2(I) (Karimi et al., 2002). Structure of the cloned construct was verified by a combination of restriction enzyme digestions as well as by sequencing analyses. Subsequently, the pIR-CMVNCR construct was transferred to Agrobacterium tumefaciens strain GV3101 by freeze-thaw transformation method (An et al., 1985). The empty pK7GWIWG2(I) vector was also introduced into A. tumefaciens strain GV3101 to use as a negative control.
Agoiniltration and virus inoculation
Leaves of N. benthamiana were infiltrated with A. tumefaciens as described previously (Canto et al., 2002). A single colony of A. tumefaciens strain GV3101 containing the pIR-CMVNCR was inoculated to LB media supplemented with 10 mM MES (pH 5.6), 20 μM acetosyringone (Fluka, USA) and antibiotics (rifampicin 10 μg/ml, ampicillin 50 μg/ml and kanamycin 50 μg/ml). The cells were grown at 28°C overnight. The cells of the overnight culture were collected by centrifugation, resuspended in infiltration buffer (10 mM MgCl2, 10 mM MES (pH5.6), and 150 μM acetosyringone) for a final OD600 of 0.5. A. tumefaciens culture was incubated at room temperature for at least 2 hour, and then infiltrated into leaves of N. benthamiana using a 1-ml disposable syringe without a needle. CMVFny was maintained in N. benthamiana by mechanical inoculation. At 3 days after agroinfiltration, the infiltrated leaves of N. benthamiana plants were mechanically inoculated with CMV-Fny by gently rubbing the leaf surface with the inoculum using Carborundum® (Fisher Scientific, USA). To determine if CMV was systemically moved in the infiltrated plants, Chenopodium quinoa that is a good local host for CMV was mechanically inoculated with sap inoculum prepared from upper leaves of the infiltrated plants at 14 days post-inoculation (dpi). The inoculated plants were kept in a growth room with a 16 h light and 8 h dark cycle at 25°C, and observation of CMV symptoms was done until 28 dpi.
Analysis of viral RNA in plants.
Total RNA was extracted from inoculated leaves at 3 days after CMV inoculation and from upper leaves at 7 dpi, using Trizol reagent and phenol/chloroform (25:24, v/ v), as described previously (Choi et al., 2011). Further precipitation, purification and DNase I treatment were performed as standard protocols (Sambrook et al., 1989). RNA concentrations were measured photometrically with a NanoDrop (Thermo Scientific, USA) and RNA quality was analyzed by 1.2 % agarose-formaldehyde gels electrophoresis. To detect CP gene of CMV-Fny from the infiltrated and the upper leaves of N. benthamiana, RT-PCR analysis was carried out using primers specific to CMV, as described previously (Choi et al., 1999). Primers specific to elongation factor 1α (Nb-EF1α) gene of N. benthamiana were used as a housekeeping gene (Choi et al., 2011). The amplified RT-PCR product was separated on 1.2% agarose gel and stained in edithium bromide (EtBr) solution. Low molecular weight (LMW) RNA was isolated from after precipitation of 200 μg total RNA with polyethylene glycol (molecular weight 8000) to 5% as well as with NaCl to 0.5 M, as described previously (Hamilton and Baulcombe, 1999). The LMW RNA (2 μg) was separated by electrophoresis in a 15 % polyacrylamide gel containing 8 M urea and 1X TBE buffer. A visualization of the 5S RNA/tRNA bands by EtBr staining was used to monitor loading of RNA samples. The RNA in gels was transferred to Hybond-N+ membrane by electronic transfer according to standard procedures and then RNA was fixed by UV crosslinking, according to manufacturer’s instructions (Bio-Rad, USA). The blot was prehybridized and hybridized using UltraHyb® buffer at 42°C for 16 hr, according to manufacturer’s instructions (Ambion, USA). The siRNAs derived from CMV genomic RNAs were detected with a digoxigenin (DIG)-labeled riboprobe probe complementary to the CP gene of CMV-Fny using T7 RNA polymerase. Blots were washed twice with 2X SSC buffer containing 0.1% SDS and with 0.5X SSC buffer containing 0.1% SDS, respectively. Signals were detected using DIG luminescent detection kit with CSPD, according to manufacturer’s instructions (Roche, USA).
RESULTS
Rationale and design of the pIR-CMVNCR construct
Sequence alignments among the members of subgroup IA and subgroup IB revealed several conserved sequences in 3′ NCRs of CMV genomic RNAs (Fig. 1). Briefly, the nt sequences of 3′ NCRs of RNA1, RNA2 and RNA3 in all CMV isolates showed over 90% sequence identity. Based on these results, we selected two highly conserved regions in the 3′ NCR of CMV RNA3 (Fig. 1) and designed a pair of primers to amplify 162bp cDNA fragment of CMV RNA3. Then, 162bp cDNA fragments were inserted into two sites of p pK7GWIWG2(I) vector, resulting in pIRCMVNCR containing 644 bp intron derived from A. thaliana (Fig. 2). This construct is expected to produce a selfcomplementary hpRNA molecule in plants and is under transcriptional control of the 35S promoter of Cauliflower mosaic virus (CaMV) and 3′ termination sequences of CaMV 35S (Fig. 2).
Transient expression of hairpin RNA interferes with virus infection
To determine whether transient expression of pIRCMVNCR could trigger an antiviral response in plants against CMV, two fully expanded leaves of N. benthamiana were infiltrated with A. tumefaciens culture carrying pIR-CMVNCR or the empty vector, respectively. At 3 dpi, N. benthamiana infiltrated with A. tumefaciens cultures carrying IR-CMVNCR was mechanically inoculated with sap inoculum of CMV-Fny. As negative controls, at 3dpi, N. benthamiana infiltrated with A. tumefaciens cultures carrying pIR-CMVNCR were mechanically inoculated with buffer alone (hereafter Cont1-plants) and N. benthamiana infiltrated with A. tumefaciens cultures carrying the empty vector were mechanically inoculated with sap inoculum of CMV-Fny (hereafter Cont2-plants).
In several independent experiments, Cont2-plants that had been infiltrated with A. tumefaciens harboring the empty vector showed wilt symptoms on the infiltrated leaves and mosaic symptoms in upper leaves at 7 dpi. In contrast, all inoculated plants that had been infiltrated with pIR-CMVNCR, except a plant remained symptomless throughout the entire testing period, similar to Cont2- plants inoculated with buffer alone (Table 1 and Fig. 3).
RT-PCR analysis confirmed this observation in upper leaves at 14 dpi. CMV RNA3 was not detectable in plants agroinfiltrated with pIR-CMVNCR construct and in Cont1-plants (named NCR in Fig. 4), whereas cDNA fragment corresponding to sequences of CMV CP gene was synthesized from the Cont2-plants (named wt-Sam in Fig. 4). The amplified cDNA fragments of the Nb-EF1α were synthesized from all samples, Cont1-plants, and Cont2-plants (Fig. 4). These results indicate CMV-genomic RNAs were specifically degraded by transient expression of pIRCMVNCR in the plants, suggesting that interference with CMV infection conferred by transient expression of pIRCMVNCR is due to the activation of RNA silencing.
Moreover, Chenopodium quinoa, a local host for CMV, was inoculated with the upper leaves of N. benthamiana agroinfiltrated with pIR-CMVNCR, resulting in none of local lesions. This result suggests that biologically active CMV-Fny was absent in the the upper leaves of N. benthamiana agroinfiltrated with pIR-CMVNCR. In contrast, C. quinoa plants inoculated with a sap inoculum from the upper leaves of Cont2-plants produced local lesions at 7 dpi (data not shown). These results suggest that the resistance response of plants elicited by hpRNA sequences in pIR-CMVNCR have detrimental effect on the replication process of CMV. It is worthwhile to mention that the interference with CMV infection triggered by the pIRCMVNCR expression was sequence-specific, because pIR-CMVNCR expression in N. benthamiana had no effects on timing and severity of Tobacco mosaic virus (TMV strain U1) symptoms in upper leaves. TMV accumulation in upper leaves of the plants agroinfiltrated with the IR-CMVNCR was similar to TMV accumulation in the plants agroinfiltrated with the empty vector or wild-type N. benthamiana plants (data not shown). Taken together, these observations support that CMV genomic RNAs in N. benthamiana expressed transiently pIR-CMVNCR were specifically degraded by the activation of RNA silencing (Fig. 2 and Fig. 4).
To determine whether the timing of transient pIRCMVNCR expression is crucial for inhibition of CMV infection, N. benthamiana was first inoculated with CMV. After 24 hr, A. tumefaciens harboring pIR-CMVNCR or the empty vector was infiltrated into the CMV-inoculated leaves or the upper leaves. In both cases, all the plants displayed typical CMV symptoms at 7 dpi, and CMV CP gene was detected from upper leaves of all sample plants (data not shown). Therefore, transient expression of pIRCMVNCR prior to virus inoculation is required to interfere with CMV infection.
Since homology-dependent resistance to virus infection is a characteristic feature of RNA silencing, we further tested whether interference with CMV infection conferred by transient expression of pIR-CMVNCR could be attributable to RNA silencing. The LMW RNA fraction was extracted from CMV-inoculated leaves that had been agroinfiltrated with IR-CMVNCR and from CMV-inoculated leaves of Cont2-plants. The buffer-inoculated leaves of Cont1-plants were used as negative controls. It is worthwhile to mention that RNA probe from CMV CP gene is not able to hybridize the sequences of the inserted cDNA of CMV 3'-NCR in pIR-CMVNCR. Thus, the probe enables us to exclude siRNA species generated from expression of pIR-CMVNCR transgene transferred by A. tumefacience. As shown in Fig. 5, 21-23 nucleotides siRNA species that are a hallmark of RNA silencing were detected in plants that had been infiltrated with pIRCMVNCR but the siRNAs were absent in Contl-plants and Cont2-plants. These results clearly show that the activation of RNA silencing plays a crucial role in the interference with CMV infection in leaves expressed transiently pIRCMVNCR.
DISCUSSION
We showed that A. tumefaciens-mediated transient expression of a homologous hpRNA results in resistance to CMV infection in N. benthaminana. Numerous studies have indicated that IR constructs of transgenes can effectively induce RNA silencing to trigger knockout of specific gene expressions (Ali et al., 2010; Allen et al., 2004; Gavilano et al., 2006; Johansen and Carrington, 2001; Kusaba et al., 2003; Liu et al., 2002; Meli et al., 2010; Pandolfini et al., 2003; Segal et al., 2003; Smith et al., 2000; Xiong et al., 2005). In addition, a number of studies showed that IR constructs of transgenes containing virusderived sequences transformed to plants can effectively protect plants from challenging viruses (Abhary et al., 2006; Bucher et al., 2006; Chen et al., 2004; Di Nicola- Negri et al., 2005; Hammond et al., 2006; Lennefors et al., 2006; Pooggin et al., 2003; Simón-Mateo and García, 2011; Tenllado et al., 2003; Tenllado et al., 2004; Vanitharani et al., 2003; Wang et al., 2000). In particular, transient expression of a hpRNA homologous to sequences of Pepper mild mottle virus (PMMoV) caused specific inhibition of PMMoV accumulation (Tenllado et al., 2003). A. tumefaciens- mediated transient expression of a hpRNA derived from the TGBp1 of potato virus X (PVX) induced RNA silencing of the triple gene block (TGB) p1 gene and resulted in the interference of PVX infection (Takahashi et al., 2006). Transient expression of a CP hairpin RNA also induced interference of PVX. The TGBp1 hpRNA showed more efficient interference of PVX infection than the CP hpRNA, but the interference was induced in the infiltrated leaves but not in the upper non-infiltrated leaves. A. tumefaciens- medicated transiently expression of a hpRNA homologous to the CP gene of potato virus Y (PVY) showed a complete and specific interference with aphid transmission of PVY in leaf tissues of N. benthamiana (Vargas et al., 2008).
Although CMV 2b protein as a viral silencing suppressor involved in binding to siRNAs in a manner analogous to p19 of tombusviruses has been directly identified (Palukaitis and García-Arenal, 2003), transient expression of pIR-CMVNCR resulted in high induction of RNA silencing in the infiltrated N. benthamiana (Fig. 4 and Fig. 5). It is plausible that transient expression of pIR-CMVNCR efficiently interferes with replication process of CMV positive- sense RNA synthesis, because 3' NCR in positivesense single strand RNA viruses contains promoter sequences that are crucial for the synthesis of positivesense strands equivalent messenger RNAs. In particular, the inhibition of replication process of CMV RNA1 and RNA2 significantly caused poor translation of the CMV replicase complex encoding RNA-dependent RNA polymerase, helicase and methyltransferase. Considering that hpRNAs less than 100 nucleotides in length are usually efficient in eliciting RNA silencing (Wesley et al., 2001) and that viral RNA genomes are usually several kb long, it would feasible to design several different hpRNAs from the sequences of CMV genomic RNAs. RNA silencing is an adaptive mechanism of defense and the viral genome is both a target for RNA degradation and a template for amplification of RNA silencing. Thus, resistance results from the interplay of host mechanisms and virus inoculum and replication. Damage and economic losses in pepper production are mainly due to CMV infection in Korea (Cho et al., 2007; Choi et al., 2001; Choi et al., 2005). We expect that the pIR-CMVNCR construct can be used to produce transgenic pepper that is highly resistant to CMV. These transgenic pepper plants will contribute to reducing damage from the virus.
적 요
Bromoviridae과 Cucumovirus속에 속하는 대표 바이러스인 오이모자이크바이러스 (Cucumber mosaic virus: CMV)는 많 은 경제적으로 중요한 원예작물 및 관상식물들에 심한 손실을 초래하는 바이러스이다. 다중염기서열 비교는 현재까지 서브 그룹1에 속하는 모든 CMV 계통들에서 3'말단부의 보전적 염 기서열들이 존재하고 있음을 보여주었다. 이런 관찰에 기초하 여, 우리는 CMV RNA3와 상동성을 가지는 162 bp 상보적 DNA를 포함하며 CMV감염에 대하여 식물 유래 RNA간섭 현 상을 유도할 수 있는 도치된 반복 구조를 가지는 머리핀 RNA (pIR-CMVNCR)를 발현시킬 수 있는 벡터를 제작하였 다. 아그로박테리움을 이용하여 IR-CMVNCR의 일시 발현은 CMV 감염을 저해하였으며, 아그로박테리움이 접종된 식물의 상엽에서는 CMV 병징이 발현되지 않았다. 또한 RT-PCR결과 는 아그로박테리움이 접종된 식물의 접종엽 및 상엽에서 CMV 서브유전자 4를 포함하는 CMV RNA들이 모두 검출이 되지 않았다. CMV 유래 작은 저해 RNA들의 축적이 관찰되 었으며, 이의 결과의 의미는 아그로박테리움에 의해 IRCMVNCR을 일시 발현시킨 야생담배 (Nicotiana benthamiana) 의 접종엽에서 RNA 간섭 현상이 유도되어 CMV 감염을 억 제시키는 것으로 판명되었다.
