The NIa-Protease Protein Encoded by the Pepper Mottle Virus Is a Pathogenicity Determinant and Releases DNA Methylation of Nicotiana benthamiana

It is well documented that the canonical function of NIa-protease (NIa-Pro) of the potyviruses is responsible for cleaving the viral polyprotein into functional proteins. Although NIa-Pro is vital for the infection cycle of potyviruses, the function of NIa-Pro in the interaction of the potyvirus host is not clear. In this study, NIa-Pro is ectopically expressed from a potato virus X (PVX) vector and infiltrates Nicotiana benthamiana wild type and 16-TGS. The pathogenicity and inhibition of host transcriptional gene silencing (TGS) are characterized. Ectopic expression of NIa-Pro from a PVX vector resulted in severe mosaic symptoms followed by a hypersensitive-like response in N. benthamiana. Furthermore, PepMoV NIa-Pro was able to reverse established TGS of a green fluorescent protein transgene by reducing methylation of promoter sequences in N. benthamiana and possessed the capacity to interfere with the global methylation of N. benthamiana. Taken together, the results of this study likely suggest that PepMoV NIa-Pro is a pathogenicity determinant and a potent suppressor of host TGS and suggest that NIa-Pro may employ novel mechanisms to suppress host antiviral defenses. To the best of our knowledge, this is the first report of a plant RNA virus modulating host TGS in a novel manner by interfering with the establishment of the methylation step of the plant DNA methylation pathway.

It is well documented that the canonical function of NIa-protease (NIa-Pro) of the potyviruses is responsible for cleaving the viral polyprotein into functional proteins. Although NIa-Pro is vital for the infection cycle of potyviruses, the function of NIa-Pro in the interaction of the potyvirus host is not clear. In this study, NIa-Pro is ectopically expressed from a potato virus X (PVX) vector and infiltrates Nicotiana benthamiana wild type and 16-TGS. The pathogenicity and inhibition of host transcriptional gene silencing (TGS) are characterized. Ectopic expression of NIa-Pro from a PVX vector resulted in severe mosaic symptoms followed by a hypersensitive-like response in N. benthamiana. Furthermore, PepMoV NIa-Pro was able to reverse established TGS of a green fluorescent protein transgene by reducing methylation of promoter sequences in N. benthamiana and possessed the capacity to interfere with the global methylation of N. benthamiana. Taken together, the results of this study likely suggest that PepMoV NIa-Pro is a pathogenicity determinant and a potent suppressor of host TGS and suggest that NIa-Pro may employ novel mechanisms to suppress host antiviral defenses. To the best of our knowledge, this is the first report of a plant RNA virus modulating host TGS in a novel manner by interfering with the establishment of the methylation step of the plant DNA methylation pathway.

INTRODUCTION
Potyviruses (genus Potyvirus, family Potyviridae) constitute the largest group of plant-infecting RNA viruses affecting major crops worldwide. Potyviruses encode a large polyprotein that is subsequently cleaved by three virus-encoded proteinases to yield up to 10 mature proteins (Shukla et al., 1994). Additionally, a P3N-PIPO is a fusion of the N-terminal part of P3 and the PIPO protein-coding regions by polymerase slippage (Rodamilans et al., 2015;Untiveros et al., 2016).
The main viral proteinase, NIa, is a polyprotein consisting of two domains, viral protein genome-linked (VPg) and protease (NIa-Pro), separated by an inefficiently utilized self-proteolytic site. NIa-Pro encoded by potyviruses plays many roles during the virus infecting cycle. The canonical function of NIa-Pro is the main protease of the virus responsible for the cleavage of most sites in the polyprotein (Sun et al., 2010). Tobacco etch potyvirus encoding NIa-Pro can bind NIb, which is critical for tobacco etch virus (TEV) genome replication (Daros et al., 1999). Further proteomic studies demonstrated that TEV NIa-Pro could interact with 76 host proteins involved in response to host stress, metabolism, and photosynthesis (Martínez et al., 2016). Additionally, NIa-Pro induced vein necrosis in tobacco by potato virus Y (PVY) isolates (Faurez et al., 2012), which indicated that NIa-Pro of the potyviruses may be a pathogenicity determinant.
Cytosine DNA methylation, one of the fundamental aspects of stable epigenetic modification, modulates coding genes that involve various cellular biological processes and molecular functions to counter virus infection (Wang et al., 2019). The cytosines in the genome of DNA viruses could be directly methylated through plant DNA methylation; for example, geminivirus transcription is usually disrupted by host DNA methylation, which occurs in the viral genome through RNAdirected DNA methylation (RdDM) (Coursey et al., 2018). Although RNA viruses cannot be directly methylated by plant DNA methylation owing to the absence of DNA in the replication process, the 2b protein encoded by cucumber mosaic virus (CMV) could bind the AGO4-related small RNAs and impairs AGO4 activities to release the DNA methylation of Nicotiana benthamiana, suggesting the modulation of RNA virus replication by plant DNA methylation (Hamera et al., 2012).
To counter the suppression of viral transcription mediated by plant DNA methylation, DNA viruses, mainly from geminiviruses, encode DNA methylation suppressor proteins. The C2 and C4 proteins encoded by tomato leaf curl Yunnan virus (TLCYnV), the V2 protein encoded by Tomato yellow leaf curl virus (TYLCV), and the βC1 and the Rep proteins encoded by geminiviruses can inhibit host transcriptional gene silencing (TGS) and modulate the plant DNA methylation pathway (Yang et al., 2011;Zhang et al., 2011;Rodrıguez-Negrete et al., 2013;Xie et al., 2013;Wang et al., 2014). However, the DNA methylation suppressor proteins encoded by plant RNA viruses have little been documented until now (Hamera et al., 2012).
PepMoV was first identified as an atypical species of potyvirus (Nelson and Wheeler, 1972) and is known to cause serious symptoms in many solanaceous plants (Han et al., 2006). Although most proteins encoded by potyviruses have been well characterized, the functions of PepMoV-encoded proteins are mostly annotated through comparison with other species of potyviruses and remain to be identified.
In this study, the functions of PepMoV-encoded NIa-Pro were characterized by ectopic expression of NIa-Pro from a potato virus X (PVX) vector, and the results demonstrated that PepMoV NIa-Pro is a pathogenicity determinant and a potent suppressor of host TGS and suggest that NIa-Pro may employ novel mechanisms to suppress host antiviral defenses.

Plant Materials
Wild-type and 16-TGS lines (Buchmann et al., 2009) of Nicotiana benthamiana Domin were used in this study. Plants were grown inside a culture room set at 25 ± 2 • C and 60% relative humidity with a 16-h light/8-h dark illumination. The third or fourth leaves of 6-to 8-week-old plants were infiltrated individually with one of the agrobacterium cultures containing the plasmid made in this study.

Plasmid Constructs and Agroinfiltration
The NIa-Pro gene sequence was cloned from PepMoV isolate HN (GenBank accession No. LN832375) (Zhang et al., 2019) and inserted into a PVX-based vector (pGR106) between the ClaI and SalI sites. The recombinant pGR106 vectors expressing the NIa-Pro gene or βC1 gene [as a positive control, cloned from tomato yellow leaf curl China virus betasatellite (TYLCCNB) and described previously] (Buchmann et al., 2009) were transformed into Agrobacterium tumefaciens GV3101. The Agrobacterium cultures were cultured in Luria broth with appropriate antibiotic selection at 28 • C for 16 h. The cells were centrifugated and resuspended in infiltration medium (10 mM of MgCl 2 and 20 mM of acetosyringone) to an OD600 0.4 to 0.6 and incubated at room temperature for 3 h before injection into leaves. The underside of the third or fourth leaf of N. benthamiana was infiltrated with syringe (Rajamaki and Valkonen, 2009).

H 2 O 2 Detection
The production and accumulation of H 2 O 2 was detected visually in leaves using the 3,3 -diaminobenzidine (DAB)-HCl (Sigma) uptake method as described previously (Sharma et al., 2010).

DNA Methylation Analysis Using Restriction Digestion and Quantitative Real-Time PCR
Genomic DNA was extracted from plant leaves using the cetyl trimethylammonium bromide (CTAB) reagent (Tiangen, Beijing, China). DraI (TaKaRa, Kyoto, Japan), a methylationinsensitive restriction endonuclease, and McrBC (New England Biolabs, Ipswich, MA, United States). A methylation-dependent restriction endonuclease was used for digestion analysis of genomic DNA according to the manufacturer's instructions. Briefly, genomic DNA (10 ng) was digested by 50 U of the respective enzyme in a 50-ml reaction mix. Digested genomic DNA was immediately separated by 1.5% agarose gel electrophoresis. For quantitative real-time PCR (qPCR) analysis, specific primer pairs for 35S promoter or GFP gene (Supplementary Table S1) were used, and NbActin was set as reference gene (Li et al., 2015).

RNA Extraction and Reverse Transcription Quantitative Real-Time PCR Analysis
Total RNA was isolated from infiltrated leaves of wild-type or 16-TGS N. benthamiana using TRIzol reagent (Tiangen, Beijing,  Table  S1) were designed by Primer Premier 5 software or referred to published documents, and NbGADPH was set as reference gene .

Western Blotting
Total protein was extracted from infiltrated systemic leaves as described previously . Western blotting was performed with primary mouse monoclonal antibody [green fluorescent protein (GFP) or PVX CP protein monoclonal antibody, HuaAn Beijing, China], followed by goat antimouse secondary antibody conjugated to alkaline peroxidase (Bio-Rad). Blotted membranes were washed thoroughly and visualized using NBT/BCIP.

Green Fluorescent Protein Images
To visualize GFP in TGS assays, representative plants were photographed at 14 dpi with a Canon 400D digital camera with a 58-mm yellow filter. GFP fluorescence was activated by a hand-held longwave UV lamp (100 W, Black Ray model B 100 A; UV products).

Phylogenetic Analyses of PepMoV-Encoded NIa-Pro
To examine the evolutionary relationships of NIa-Pro sequences from different potyviruses, a phylogenetic dendrogram of the complete NIa-Pro nucleotide sequences of 16 representative potyvirus species was generated. As shown in Figure 1A, two major subclusters are grouped. Subcluster I contains 11 of 16 potyviruses, and cluster II consists of five of 16 potyviruses. The PepMoV NIa-Pro protein is located in cluster II and shares the highest sequence identity with PVY. Sequence alignment showed that potyvirus NIa-Pro proteins are highly divergent in amino acid sequences. The structure of the NIa-Pro protein was analyzed by the Conserved Domain Database (CCD) available at the National Center for Biotechnology Information (Marchler-Bauer et al., 2011), and only one conserved domain architecture was retrieved for the NIa-Pro protein ( Figure 1B). All NIa-Pro proteins from potyviruses contain this conserved domain. The presence of the same conserved position in many potyviruses suggests that NIa-Pro may play an important role during virus infection.

NIa-Pro Is a Pathogenicity Determinant in Nicotiana benthamiana
To determine whether ectopically expressed NIa-Pro is a symptom determinant in Nicotiana benthamiana, the NIa-Pro gene was inserted into a PVX-based vector (pGR106) for overexpression. As shown in Figure 2 (upper panel), the phenotype in plants infected with PVX was similar to that of plants infected with PVX-NIa-Pro at 10 dpi. At 20 dpi, PVXinfected plants showed a loss of veinal chlorosis and mild mosaic symptoms; however, continuous NIa-Pro expression resulted in symptoms of severe mosaic, chlorosis, and leaf deformity at 20 dpi (Figure 2, lower panel). In addition, pronounced necrotic symptoms were observed on the systemic leaves of PVX-βC1infected plants, but plants inoculated with PVX-NIa-Pro failed to develop the necrotic phenotype (Figure 2, lower panel).

NIa-Pro Elicits a Hypersensitive-Like Response in Nicotiana benthamiana
To determine whether NIa-Pro elicits the hypersensitive-like response (HLR), PVX-, and PVX-NIa-Pro-infected leaves were analyzed by the DAB uptake method. DAB polymers can stain the accumulated H 2 O 2 as a deep brown color, which can be visualized after ethanol clearing of the tissue. As anticipated, compared with PVX-infected plants, PVX-NIa-Pro-infected plants accumulated high concentrations of H 2 O 2 in systemically infected leaves ( Figure 3A). To determine whether severe symptoms in the presence of NIa-Pro are a consequence of higher accumulation of PVX, RT-qPCR and Western blotting against PVX CP protein were employed to examine PVX genomic RNA levels and accumulation of CP protein at 20 dpi (Figures 3B,C). More PVX genomic RNAs and CP proteins accumulated in PVX-NIa-Pro-infected plants compared with PVX-infected plants at 20 dpi, suggesting that NIa-Pro is a virulence factor that enhances replication of PVX in N. benthamiana. These data reveal that NIa-Pro is a symptom determinant that can elicit an HLR when expressed from PVX-NIa-Pro.
To test if there was any association between NIa-Pro and endoplasmic reticulum (ER) stress, the expression levels of key genes involved in ER stress, including Bip (ER-localized binding protein), HSP90-2 (heat shock protein 90-2), and bZIP60 (basic leucine zipper protein 60), were quantified. qRT-PCR analysis showed that the expression of Bip and HSP90-2 was significantly induced by NIa-Pro; otherwise, the expression of bZIP60 was not influenced by NIa-Pro (Figure 4). These results suggest that NIa-Pro can induce ER stress by stabilizing proteins for correct folding or re-folding in response to ER stress but not the unfold protein response (UPR) to ER stress.

NIa-Pro Reverses Established Methylation-Mediated Transcriptional Gene Silencing
Plant DNA methylation modulates coding-gene expression to induce TGS (Law and Jacobsen, 2010). To determine whether NIa-Pro inhibited plant methylation-mediated TGS, transgenic N. benthamiana 16-TGS plants containing a GFP transgene downstream of a transcriptionally silenced CaMV 35S promoter were employed in this study. Previously, this assay demonstrated the TGS suppression activity of several geminivirus-encoded proteins (Raja et al., 2008;Yang et al., 2011). The mock culture (GV3101 Agrobacterium tumefaciens cultures), PVX (negative control), PVX-βC1 (positive control), or PVX-NIa-Pro 16-TGS plants were inoculated in the third and fourth leaves. By 14 dpi, mild symptoms were exhibited in systemic leaves of PVXinfected plants; however, the typical βC1-associated phenotypes, including enations and upward leaf curling, developed in systemic leaves of PVX-βC1-infected plants. The phenotypes induced by PVX and PCX-βC1 were in agreement with those described previously (Yang et al., 2011). Severe mosaic symptoms developed in plants infected with PVX-NIa-Pro. In addition, yellow-green GFP fluorescence was visible in plants infected with PVX-βC1 and PVX-NIa-Pro ( Figure 5A). As anticipated, Western blot analysis verified the visible fluorescence in PVX-NIa-Pro-or PVX-βC1-inoculated plants ( Figure 5B).
To determine whether NIa-Pro affects global DNA methylation, a methylation-dependent restriction enzyme (McrBC), which only cleaves methylated DNA sequences (Stewart et al., 2000), was used. Genomic DNA was extracted from leaves of N. benthamiana plants inoculated with mock culture (GV3101 A. tumefaciens cultures), PVX (negative control), PVX-βC1 (positive control), or PVX-NIa-Pro and subjected to restriction digestion. A mock digestion with no  Values are the mean ± SD. Highly significant differences (**p < 0.01) between samples in each pair are indicated. enzyme was included (sham) as a control. All four genomic DNA samples were able to digest by DraI, a methylation-independent restriction endonuclease, as indicated by the presence of a "smear" of different sized fragments on the gel (Figure 5C). However, although McrBC was able to cleave a significant portion of the genomic DNA of plants with Mock and PVX into small fragments, DNA from plants expressing NIa-Pro and βC1 was largely resistant to McrBC digestion ( Figure 5C). DNA quantification by qPCR of 35S promoter and adjacent GFP gene has shown that the DNA level of 35S of PVX-NIa-Pro was relative higher than that of mock and PVX; otherwise, no difference appeared in adjacent GFP gene ( Figure 5D). Thus, the expression of NIa-Pro in N. benthamiana plants likely reduced DNA methylation at a genome-wide level.

N. benthamiana plants expressing empty vector (PVX) and
NIa-Pro were analyzed at 15 dpi. As shown in Figure 6, overexpression of NIa-Pro significantly induced NbAGO4 expression (32% of that of control plants; p < 0.01), and expression of NbMET1, NbDRM2, and NbCMT3 increased 46% (p < 0.01), 38% (p < 0.01), and 26% (p < 0.05), respectively (Figure 6). NbROS1 and NbROS2 expression increased by 18 and 15%, respectively, but not significantly (Figure 6). Taken together, these results demonstrate that the expression of NIa-Pro significantly suppresses the expression of NbAGO4, a gene responsible for establishing DNA methylation, which is the first step of the plant DNA methylation pathway. In addition, activation of NbROS1 and NbROS2, two DNA glycosylases and repressors of TGS, may also contribute to TGS suppression by NIa-Pro. Values are the mean ± SD. Highly significant differences (**p < 0.01) between samples in each pair are indicated.

DISCUSSION
PepMoV was first identified as an atypical isolate of the genus Potyvirus for serum differentiation and low amino acid sequence identity (Zitter, 1972). For example, PepMoV NIa-Pro exhibits 63.5% identity in sequence with PVY NIa-Pro and only 45.9% identity with TEV NIa-Pro (Mestre et al., 2000). Although the canonical function of NIa-protease (NIa-Pro) encoded by all potyviruses is responsible for cleaving the viral polyprotein into functional proteins, the other functions of NIa-Pro in the virus infecting cycle of potyviruses are inconsistent. TEV NIa-Pro could not induce a hypersensitive response (HR); however, PVY NIa-Pro could induce an HR in Ry-mediated resistance potatoes (Mestre et al., 2000). PepMoV NIa-Pro shared the highest nucleotide and amino acid sequences with PVY (Figure 1), and PVY NIa-Pro induced tobacco vein necrosis with HR (Faurez et al., 2012), suggesting PepMoV NIa-Pro would be a pathogenicity determinant. In this study, by using a PVX vector as an ectopic expression system, we found that PepMoV NIa-Pro could result in severe mosaic symptoms (Figure 2) and enhanced PVX replication (Figures 3B,C) but was not able to induce vein necrosis in leaves in Nicotiana benthamiana, which demonstrated that PepMoV NIa-Pro is a symptom determinant; however, it functioned in a different manner with PVY NIa-Pro.
Plant responses to infection by pathogens are often accompanied by a HLR, which is associated with an oxidative burst (including O 2 and H 2 O 2 ) (Bao and Howell, 2017). PepMoV NIa-Pro can induce H 2 O 2 accumulation during infection with a PVX-based vector expressing NIa-Pro ( Figure 3A). Additionally, PepMoV NIa-Pro could significantly activate genes expressing Each mean value was derived from nine plants in three independent experiments. Student's t-test was performed to assess differences between PVX-NIa-and PVX vector-infected plants. Values are the mean ± SD. Significant differences (*p < 0.05) and highly significant differences (**p < 0.01) between samples in each pair are indicated.
Bip and HSP90-2 but did not change the expression of bZIP60 (Figure 4). Bip is a chaperone gene that restores proper protein folding during ER stress (Bao and Howell, 2017), the HSP90-2 gene is involved in stabilizing proteins for correct folding or re-folding in response to ER stress (Sangster et al., 2007), and bZIP60 is the downstream transcription factor that activates genes required for coping with UPR (Arias Gaguancela et al., 2016). Therefore, PepMoV NIa-Pro likely induces N. benthamiana ER stress with a branch pathway of protein folding but does not activate the UPR.
It was well documented that DNA viruses, such as geminivirus-encoded pathogenicity determinants, also function as TGS suppressors targeting different steps of the TGS pathway. The βC1 protein of TYLCCNB directly interacts and inhibits host S-adenosyl homocysteine hydrolase (SAHH) activity to suppress host DNA methylation (Yang et al., 2011). The C2 protein of beet curly top virus (BCTV) attenuates the degradation of S-adenosylmethionine decarboxylase 1 (SAMDC1) of Arabidopsis to inhibit host DNA methylation (Zhang et al., 2011). Intriguingly, a recent study demonstrated that the pathogenicity determinant of RNA viruses could also modulate the host methylation pathway. The 1a protein encoded by CMV, which is essential for viral infection, could recruit a novel methyltransferase from Nicotiana tabacum to methylate itself, which promotes CMV infection (Kim et al., 2008). Our results showed that PepMoV NIa-Pro can reverse the established TGS of N. benthamiana, suggesting that PepMoV NIa-Pro functioned as a host TGS suppressor. To our knowledge, this is the first report of plant RNA virus directly suppressing the plant DNA methylation pathway.
The DNA methylation in plants is not a static pathway and is dynamically modulated with three steps of establishing DNA methylation, maintenance methylation, and demethylation. Three types of cytosine methylation have been extensively characterized in plants: symmetric CG and CHG sites and asymmetric CHH (where H = A, C, or T) sites (Law and Jacobsen, 2010). Plant DNA viruses, such as geminivirus Rep protein, are the main viral proteins that can alter the levels of expression of the plant DNA methyltransferases responsible for the maintenance of symmetric methylation (Rodrıguez-Negrete et al., 2013), and AC5 protein of mungbean yellow mosaic India virus (MYMIV) inhibits methylation of cytosine in CHH and CHG to suppress NbDRM2 expression (Li et al., 2015). These results suggest that plant DNA viruses can modulate the host DNA methylation pathway in different manners. However, evidence of plant RNA viruses interacting with the host DNA methylation pathway was observed in a study where the CMV 1a protein was methylated by a novel protein involving host DNA methylation (Kim et al., 2008); this is not direct evidence of plant RNA viruses with the capacity to modulate host DNA methylation. PepMoV NIa-Pro can reverse established TGS by suppressing NbAGO4 (Figure 6), which is critical for establishing DNA methylation in the first step of the plant DNA methylation pathway (Law and Jacobsen, 2010). NIa-pro significantly activates the expression of NbMET1, NbDRM2, and NbCMT, which catalyze CG, CHG, and CHH methylation (Law and Jacobsen, 2010), suggesting that NIa-Pro likely activates the second step of plant DNA methylation. Interestingly, NIa-Pro could also activate the expression of NbROS1 and NbROS2 (Figure 6), which has been demonstrated to demethylate DNA methylation (Kim et al., 2019). It is likely that PepMoV NIa-Pro also possesses the capacity to inhibit the plant DNA methylation pathway through activating demethylation, which is the last step of plant DNA methylation. Further studies of the molecular mechanisms of PepMoV NIa-Pro modulating the plant DNA methylation pathway would contribute to uncovering novel modulating mechanisms of plant DNA methylation by plant RNA viruses.
Overall, the observations of this study show that PepMoV NIa-Pro is a multifunctional protein, and to the best of our knowledge, this is the first report testifying that a potyvirus NIa-Pro protein can both determine the symptom and induce an HR and modulate plant DNA methylation in multiple manners.

CONCLUSION
In conclusion, the results of this study suggest that PepMoV NIa-Pro is a pathogenicity determinant and a potent suppressor of host TGS and suggest that NIa-Pro may employ novel mechanisms to suppress host antiviral defenses.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any qualified researcher.

AUTHOR CONTRIBUTIONS
S-BZ, X-WL, and YL designed the research and prepared the draft of the manuscript. Y-NG, R-QT, YZ, JP, and OX performed the experiments. Z-HZ, D-YZ, and HL collected the data. All authors reviewed and approved the final version of the manuscript.