RNA silencing-related genes contribute to tolerance of infection with potato virus X and Y in a susceptible tomato plant

Wild-type tomato (Solanum lycopersicum cv. Moneymaker) and transgenic tomato plants created from Moneymaker were used in this study. DCL2 and DCL4- and RDR6-repressed (hpDCL2.4 and hpRDR6) tomato plants were previously created using the host RNA silencing mechanism [45, 46]. Tomato plants with AGO2a, AGO2b and AGO3 (Solyc02G069260, Solyc02G069270, and Solyc02G096280, registered in the tomato genome database at https://​solgenomics.​net/​organism/​Solanum_​lycopersicum/​genome) repressed (hpAGO2.3) were created in the same manner as hpDCL2.4 tomato plants [45]. The inverted repeat (IR) sequences were constructed by placing parts of AGO2, AGO3 (947 bp) in a head-to-head orientation across an intron sequence to create an IR sequence (Fig. 1a). The IR sequences were cloned into the BglII/KpnI site of the binary vector pIG121-Hm [45] downstream of the CaMV-35S promoter and introduced into Agrobacterium tumefaciens strain EHA105 to transform the tomato plants. T3-generation plants were used in this study. In addition, tomato plants transformed with pIG121-Hm containing an empty cassette were created and used as a negative control (empty vector). PVX derived from the infectious clone pP2C2S, which was constructed with the PVX-UK3 strain [47], as well as the N [48] and O [49] strains of PVY (PVY^N and PVY^O) were used for the inoculation tests in this study.

Tomato plants were cultivated in a growth room at 21–23 °C with a 16-h photoperiod. PVX- and PVY-infected leaf discs (50 mg) were ground in inoculation buffer (0.1 M phosphate buffer [pH 7.0], 1% 2-mercaptoethanol). The crude sap was mechanically inoculated onto the first and second leaves of 2-week old tomato plants. At the same time, other plants were inoculated with inoculation buffer alone as a control (mock inoculation). Three plants of each transgenic and wild-type were inoculated with these viruses and analyzed.

PVY genomic RNA was detected using RT-PCR as follows. Tomato leaves were ground in liquid nitrogen, and total RNA was extracted with TRIzol™ Reagent (Thermo Fisher Scientific, Inc., Waltham, MA, USA). All RNA samples were treated with RNase-free DNase I (Roche Diagnostics, Basel, Switzerland). cDNA was synthesized from 1 μg of RNA extract using the modified Moloney Murine Leukemia Virus (MMLV) reverse transcriptase ReverTra Ace® (Toyobo, Osaka, Japan). PCR was performed to detect viral genomic RNA and endogenous mRNA. The mixture (25 μl) contained cDNA corresponding to 0.1 μg RNA, 0.4 μM of each of the specific primer pairs (Additional file 1: Table S1), 0.2 mM deoxyribonucleotide triphosphate (dNTPs), and 0.625 U Ex Taq™ DNA polymerase (Takara, Otsu, Siga, Japan). PCR mixtures were incubated for 2 min at 94 °C, followed by 28 cycles at 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 60 s. The PCR products were fractionated through 1.2% agarose gel electrophoresis. Quantitative PCR (qPCR) was performed using the AriaMx real-time PCR system (Agilent Technologies, Santa Clara, CA, USA). The reaction mixture (25 μl) contained 0.3 mM (each) forward and reverse primers (Additional file 1: Table S1), 0.2 mM dNTPs, 0.625 U Ex Taq™ DNA polymerase (Takara), SYBR Green (1/800 dilution) (Thermo Fisher Scientific), and cDNA obtained by reverse transcribing 50 ng of total RNA. Samples were incubated for 2 min at 95 °C, followed by 39 cycles at 95 °C for 10 s, and at 58 °C and 72 °C for 20 s each.

Accumulation of the PVX and PVY CP and genomic RNA was investigated through western and northern blotting as previously described [50]. Upper uninoculated leaves from each plant at 15, 20, 25, and 30 days post-inoculation (dpi) were harvested and total RNA preparations (1–10 μg) were extracted using the TRIzol™ Reagent. After denaturing the RNA extracts by heating at 65 °C for 15 min in a solution containing RNA denaturation buffer (0.9 M disodium phosphate, 0.1 M monosodium phosphate, 37% formaldehyde, 0.05% formamide), the samples were loaded into 1.4% agarose gels containing 5% formaldehyde and 1 × 3-(N-morpholino) propanesulfonic acid buffer and run at 100 V for 30 min. Following transfer to a nylon membrane (Hybond-N; GE Healthcare, Chicago, IL, USA), hybridization with a digoxigenin-labeled probe (Roche Diagnostics) was performed to detect the 3′-terminal regions of PVX genome segments. Chemiluminescence signals with CDP-star (Millipore Sigma, St. Louis, MO, USA) were quantitatively detected using a LAS-4000 mini-imaging system (GE Healthcare).

Western blotting was conducted as previously described [51]. Total proteins were separated through electrophoresis in 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis Bis–Tris gels using Tris–glycine buffer, followed by electrotransfer onto a polyvinylidene difluoride membrane. For the detection of viral coat proteins, primary antibodies raised against PVX and PVY CP, provided by Japan Plant Protection Association, were used at a 1:1000 dilution, while alkaline phosphatase-conjugated goat antirabbit immunoglobulin G (Thermo Fisher Scientific) was used as the secondary antibody. Chemiluminescence signals were detected using the CDP-Star reagent in a LAS-4000 mini-imaging system.

We initially confirmed repression of the expression levels of DCL2, DCL4, AGO2, and RDR6 in hpDCL2.4, hpAGO2.3, and hpRDR6 plants through real-time RT-PCR. There was no significant difference in the expression levels of the DCL2d in hpDCL2.4 plants with those detected in empty-vector transgenic plants (Fig. 1b). In addition, the expression levels of DCL2a, b, c and DCL4 were also decreased to approximately 3%, 10%, 8% and 15%, respectively, in transgenic tomato plants compared with empty vector transformed plants. In hpAGO2.3 and hpRDR6 plants, the expression levels of both AGO2b and RDR6 genes were decreased to approximately 29% and 36% of the wild-type levels. In contrast, the expression levels of the AGO2a and b were not altered significantly in hpAGO2.3 plants. We inoculated these transgenic tomato plants with PVX and observed symptoms in non-inoculated upper leaves of all inoculated plants at 5 dpi (Fig. 2, Table 1). The symptoms that developed in hpDCL2.4 and hpAGO2.3 were more severe than those that developed in hpRDR6 and control transgenic plants with the empty vector. Specifically, necrotic spots began to develop in hpAGO2.3 plants on the second upper leaf from the inoculated leaf at 5 dpi, and the necrotic symptoms spread systemically at 20 dpi (Fig. 2). More severe dwarfing, as well as mosaics and leaf-malformation were observed in the hpDCL2.4 plants (Fig. 2). We investigated the accumulation of PVX CP and genomic RNA through western blotting and northern blotting (Fig. 3a, b) to determine whether the difference in symptoms was associated with the multiplication of PVX in these plants. The three transgenic tomato plants (i.e., hpDCL2.4, hpAGO2.3, hpRDR6) inoculated and those with the empty vector (Empty) all accumulated CP and the full-length genomic RNA comparably at 15 dpi (Fig. 3b). Of note, the accumulation of subgenomic RNAs differed among samples. The levels of DCLs, AGOs, and RDR6 mRNAs in transgenic lines infected with PVX were investigated by RT-qPCR (Fig. 3c). The levels of DCL2a, DCL2c, and DCL4 mRNAs in the hpDCL2.4 line infected with PVX were lower than those of the healthy empty vector-transformed plants. Unexpectedly, the mRNA levels of DCL2a, DCL2b, DCL2c, and DCL4 were also reduced in PVX-infected empty vector-transformed plants. In contrast, the mRNA levels of DCL2b and DCL2d were increased in PVX-infected hpDCL2.4 and empty vector-transformed plants. In hpAGO2.3 and hpRDR6 plants infected with PVX, the mRNA levels of AGO2, AGO3, and RDR6 were similar to those of PVX-infected and healthy empty vector-transformed plants.

Subsequently, we mechanically inoculated PVY^N into hpDCL2.4, hpAGO2.3, and hpRDR6 plants. Symptoms were only observed on the non-inoculated upper leaves of hpDCL2.4 (Fig. 4, Table 1), suggesting systemic infection of hpDCL2.4 plants with PVY^N. PVY^N CP and genomic RNA were detected through western blotting and RT-PCR to confirm the systemic infection of hpDCL2.4 and to examine whether the other inoculated plants had latent systemic infections of PVY^N. PVY CP was detected via western blotting only in DCL2.4 plants, which exclusively showed symptoms (Fig. 5a). PVY genomic RNA was detected through RT-PCR in all the inoculated plants at 15 dpi under 30 cycles of amplification. However, after 20 cycles of amplification, it was detected only in hpDCL2.4 plants (Fig. 5b). RT-qPCR detected a reduction of the mRNA levels of all tested genes in PVY^N-infected hpDCL2.4, hpAGO2.3, and hpRDR6 plants compared with those measures in PVY^N-infected empty vector-transformed plants. However, the mRNA levels of DCL2d, AGO2a, and RDR6 in PVY-infected empty vector-transformed plants were unexpectedly lower than those observed in PVY-infected hpDCL2.4, hpAGO2.3, and hpRDR6 plants.

We further inoculated the parental cultivar and hpDCL2.4 plants with both PVY^N and PVY^O, and compared their susceptibility to the viruses (Table 2). This experiment was performed to rule out the possibility that stresses during the production of transgenic plants affect their susceptibility to PVY^N and PVY^O in a strain-specific manner. As expected, hpDCL2.4 plants inoculated with both strains showed symptoms; however, none of the inoculated parental cultivar plants developed symptoms (Table 2). In addition, PVY CP was detected using western blotting only in hpDCL2.4 plants inoculated with both PVYs (Fig. 6) though PVY CP was observed in PVY-inoculated Moneymaker cultivar at 40dpi (Fig. 6). PVY genomic RNA was detected in all the samples from both PVY-inoculated plants through RT-PCR at 20 dpi and in the following days. However, at 15 dpi, PVY genomic RNA was only clearly detected in the samples obtained from hpDCL2.4 (Fig. 6). These results indicate that PVY latently and systemically infects the Moneymaker cultivar. Moreover, silencing of DCL2 and DCL4 increases the accumulation of PVY CP and genomic RNA, resulting in the development of symptoms.