Infection of Beet necrotic yellow vein virus with RNA4-encoded P31 specifically up-regulates pathogenesis-related protein 10 in Nicotiana benthamiana

Previous studies showed that N. benthamiana inoculated with a BNYVV isolate without RNA4 (BN3, which contains BNYVV RNAs 1, 2 and 3) does not show severe symptoms, whereas a BNYVV isolate with RNA4 (BN3 + 4, which contains BNYVV RNAs 1, 2, 3 and 4) induces a downward curling of the upper leaves at 10–12 days post inoculation (dpi) in N. benthamiana. The plant is stunted and these first leaves that undergo curling gradually wilt[13, 30]. After 15 dpi, the newly-grown leaves recovered in plants infected by BN3 without RNA4[30]. Plant development is regulated by many genes in different situations, such as biotic stresses. The relative expression levels of select genes, involved with gene silencing and pathogenesis, were determined using quantitative real-time PCR (qPCR) at 11 dpi. The primers used in detection are listed in Table 1. The genes related to gene silencing were expressed at similar levels in plants infected with BN3 and BN3 + 4 (Figure 1A), except for DCL-2, which was expressed at twice the level with BN3 + 4 than with BN3. PR-a, PR-c and NPR1-1 expression levels were not different between the test plant types, although PR-b was up-regulated less than three-fold with BN3. PR-Q and PR-10 were up-regulated by tens of folds, and PR-10 had at least a 30-fold increased expression level in BN3 + 4 compared with BN3. Thus, BN3 + 4 specifically induced PR-10 and PR-Q up-regulation in N. benthamiana, and PR-10 had the highest expression level.

To examine whether other plant viruses could induce PR-10 up-regulation, N. benthamiana was inoculated with several plant viruses, including Beet black scorch virus (BBSV), Brassica yellows virus (BrYV), Barley stripe mosaic virus (BSMV), Cucumber mosaic virus (CMV), Pea enation mosaic virus (PEMV), potato virus X (PVX), Tobacco necrosis virus (TNV) and Tobacco rattle virus (TRV). We detected PR-10 expression levels in all the tested plants when symptoms of the corresponding viruses appeared on leaves at 7–14 dpi. However, the PR-10 expression levels were not up-regulated when N. benthaminana plants were inoculated with these viruses, and the PR-10 expression level increased, at most, by three-fold. This is far less than was seen with the BN3 + 4 inoculations, indicating that the high level of PR10 up-regulation in N. benthamiana was BNYVV isolate specific (Figure 1E). Of course, it is possible that there are other plant viruses that could up-regulate PR-10 gene expression levels.

To confirm PR-10 expression characteristics during a BN3 + 4 infection, we examined the expression levels of the PR-10 gene, plant height and leaf curling at 3, 5, 7, 11 and 15 dpi. Initially, all plants were investigated using a western blot analysis with antiserum specific to the viral coat protein to confirm that plants were systemically infected by BNYVV (data not shown). The PR-10 expression level was determined by qPCR, and the expression curve was parabola shaped. At 5 dpi, the PR-10 expression level in plants infected with BN3 + 4 was slightly higher than with BN3, reaching the peak of the parabola at 7 dpi. Then, expression decreased, but was still higher than with BN3, and the level of PR-10 returned to normal by 15 dpi (Figure 1B). The leaves of plants inoculated with BN3 + 4 were slightly distorted at 5 dpi, compared with those infected with BN3, and the newly-growing leaves showed severe symptoms at 7 dpi. The plant heights showed the largest disparities by 11 dpi when the stunting of plants infected with BN3 + 4 became apparent. The neonatal leaves recovered and appeared flat at 14 dpi. After 15 dpi, newly-growing stems in plants inoculated with BN3 appeared normal (Figure 1C and D).

The open reading frame (ORF) of P31 contains five potential initiation codons. To identify the translational start codon, the P31 gene nucleotide sequence was modified to produce a series of frame-shift mutants (Figure 2A). Four frame-shift mutations (P31-2, −3, −4 and −5) were created after each initiation codon, which abolished the synthesis of P31 proteins initiated at the front AUGs. N. benthamiana was inoculated with BN3 alone, RNA4 wild type (WT) or each mutants plus BN3. The expression level of PR-10 was examined and no significant differences were observed among BN3 and the mutants (Figure 2B), suggesting that PR-10 up-regulated expression required the entire P31 protein. All plants inoculated with the mutants showed mild symptoms similar to those for BN3, contrasting with the severe symptoms induced by BN3 + 4 at 11 dpi (Figure 2C and D).

To confirm P31 expression, a GUS gene without a start codon was fused to the P31 gene in RNA4 to produce P31GUS. Frame-shift mutants were also constructed and named P31GUS-2, −3, −4 and −5. Tetragonia expansa leaves were inoculated with a mixture of WT and each mutant transcript together with BN3, and every inoculated leaf produced similar local lesions at 7 dpi[31]. Histochemical GUS staining of the leaves collected from the different treatments revealed the expression of the fused proteins. Blue spots were observed on leaves inoculated with P31GUS, P31GUS-2 and −3, but the latter two frame-shift mutants produced less and smaller spots than the WT (Figure 2E). This comparative analysis permitted us to conclude that P31 translation initiation mainly occurred at the first AUG, but the second and third AUG may weakly express an incomplete protein that could not induce PR-10 expression nor severe symptoms in N. benthamiana.

BNYVV isolates maintained by mechanical inoculation into leaves of hosts often accumulate specific forms of RNA-4 deletion mutants[32, 33]. On the basis of reported spontaneous deletion sequences, two deletion mutants were created: P31-ΔL and -ΔS. When comparing the spontaneous deletion mutants with WT, P31-ΔL showed the largest deletion region, aa 107–272, and P31-ΔS the smallest, aa 130–238, as was reported for mutants isolated from France[32‐34]. The deletion regions, nt 697–1194 and nt 768–1091, were located in the 3′-half of the P31 gene (Figure 3A). The PR-10 expression levels for BN3, P31-ΔL and -ΔS did not differ compared with WT (Figure 3B). Plants infected with the mutants had mild symptoms, such as leaf slight distortions, but little was distinguishable from the symptoms induced by BN3 (Figure 3C and D). Although the spontaneous deletion mutants did not affect the replication of BNYVV[34], they affected the expression of PR-10 and attenuated the pathogenicity of P31.

Interestingly, the sequence analysis showed that P31 was rich in Cys and compared with the spontaneous deletion mutants mentioned above, most of the Cys were located in the deletion region. To examine the impact of these cysteines on PR-10 induction, alanine (Ala) scanning (substitution of Ala for Cys) was conducted for all of the Cys (C7A, C9A, C24A, C109A, C115A, C127A, C161A, C174A, C183A, C186A, C190A, C197A, C199A, C211A, C218A, C232A and C255A). Transcripts of each mutant mixed with BN3 were separately inoculated into five unfolded leaves of N. benthamiana, and different phenotypes were observed after 11 dpi. PR-10 expression levels were analyzed, and most expression patterns were in accordance with the degree of symptoms. Compared to PR-10 expression levels in plants inoculated with BN3, six Cys mutants (C174A, C183A, C186A, C190A, C197A and C199A) induced the same expression level, six Cys mutants (C7A, C9A, C24A, C109A, C211A and C218A) induced a 4- to 7-fold increase and five (C115A, C127A, C161A, C232A and C255A) induced a 10- to 15-fold increase (Figure 4A). At the same time, the severe symptoms in six substitution mutants positioned in P31^174–199 disappeared compared with the WT. Four mutants (C7A, C9A, C161A and C211A) could still induce the down curling and distortion of upper leaves, but to a lesser degree than the WT. The remaining mutants (C24A, C109A, C115A, C127A, C218A, C232A and C255A) had the same leaf shape as WT. Only C24A, C109A, C115A, C127A, C218A, C232A and C255A mutants resulted in stunted plants, but others were similar to plants infected with BN3 (Figure 4B and C). These results indicated that some Cys in P31 were required for inducing PR-10 expression, and most of them (C174, C183, C186, C190, C197 and C199) were located in the cysteine-rich region. However, the effects of Cys on PR-10 and disease symptom expression differed.

P31 has a root-specific silencing suppression function, and recent studies showed that some viral suppressors of RNA silencing mimic the host glycine/tryptophan protein to prevent gene silencing[35‐38]. P31 contains five Trp, and there is a glycine before the first and third Trp, respectively. Thus, we used Ala to replace Trp 103, 152, 251, 259 and 276. After 12 dpi, PR-10 expression levels in the test plants infected with the mutants appeared for the same as those infected with BN3 (Figure 5A). Symptoms were not observed except for with the W152A mutant, which showed a slight distortion of the upper leaves (Figure 5B and C). Our mutation analyses revealed that every Trp in P31 was important for inducing PR-10 up-regulation and symptom development.

The PR-10 gene of N. benthamiana (NbPR10) was cloned by RT-PCR, and it was 483 bp long and capable of encoding a 160-aa peptide. There was a typical GxGGxG motif at 47–52 aa, which is known as the phosphate-binding loop (P-loop), which is frequently found in protein kinases as well as in nucleotide-binding proteins (Figure 6A)[30]. A yeast two-hybrid (YTH) assay was performed to determine whether NbPR-10 and P31 had any intrinsic ability to interact. The preliminary results indicated that they could not interact directly in yeast (Figure 6B).

DNA manipulation and cloning were performed as described by Sambrook et al.[65]. The mutants used in the experiment were produced by PCR-based, oligonucleotide-directed mutagenesis based on full-length RNA4 cDNA clone pUOF1-6[13]. Amino acid scanning mutants were also obtained by site-directed mutagenesis. All constructs were examined for sequence integrity.

T. expansa and N. benthamiana were used for viral propagation and the observation of symptom phenotypes[13, 21]. The BNVYY laboratory isolated BN3 (RNA1 + 2 + 3) was used[66]. Unless otherwise stated, virus isolates were propagated in inoculated leaves of T. expansa. Infectious cDNA clones were linearized by Xba I and transcribed at 37°C for 2 h with a T7 RNA polymerase kit as described by the manufacturer (Promega). The leaves of test plants were inoculated with the in vitro-synthesized transcripts from each cDNA clone mixed with BN3. Local lesions generally appeared at 5–8 dpi, while systemic symptoms in N. benthamiana appeared at 10–12 dpi.

Total RNA was extracted from 0.2 g of top leaves of tested N. benthamiana and used for RT-PCR detection[67]. RT-PCR was conducted using Moloney Murine Leukemia Virus Reverse Transcriptase as described by the manufacturer (Promega). cDNA was synthesized from 3 μg total RNA with oligo(dT) primer. The primers used for BNYVV detection were described previously[39]. Each experiment was repeated at least two times, and four plants were inoculated with each mutant. The total RNA extracted from each plant was tested separately by RT-PCR and qPCR.

cDNA synthesized by RT was used for qPCR. The qPCR was performed using a Bio-RAD CFX96 Touchâ„¢ Real-Time PCR Detection System according to the manufacturer’s instructions, and the primer pairs used for qPCR are found in Table 1. The level of PP2A transcript amplified with primer pair PP2A-F and PP2A-R was used as an internal control. Quantification and statistical analysis of the qPCR results were performed using the software Bio-Rad CFX Manager Version 3.1 (Bio-Rad Laboratories Inc.).

Freshly tested plant samples were incubated for 4–12 h at 37°C in the darkness in sufficient X-Gluc staining buffer (2 mM 5-bromo-4-chloro-3-indolyl β-D-glucuronic acid, 100 mM sodium phosphate buffer at pH 7.0, 10 mM EDTA, 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide and 0.1% Triton X-100). Then, chlorophyll was removed with 70% (v/v) ethanol at room temperature.

During a database (http://​solgenomics.​net/​) search with Nicotiana tabacum PR-10 (GenBank ID:JQ041907), a sequence from N. benthamiana highly homologous to NtPR-10 was identified. A pair of specific primers (PR10-forward: 5′-ATGGGTGTCACTACCTATACTCATG-3′ and PR10-reverse: 5′-TTAAACATAGAGAGAAGGATTAGCG-3′) were synthesized based on the newly identified sequence to clone the putative PR-10 gene (NbPR-10). Total RNA was isolated from young N. benthamiana leaves infected by BN3 + 4 and was reverse-transcribed to cDNA. The PCR product was gel purified and ligated into a pMD19-T vector (Takara) to obtain pMD–NbPR10 and confirmed through DNA sequencing.