Strawberry Vein Banding Virus P6 Protein Is a Translation Trans-Activator and Its Activity Can be Suppressed by FveIF3g

The strawberry vein banding virus (SVBV) open reading frame (ORF) VI encodes a P6 protein known as the RNA silencing suppressor. This protein is known to form inclusion like granules of various sizes and accumulate in both the nuclei and the cytoplasm of SVBV-infected plant cells. In this study, we have determined that the P6 protein is the only trans-activator (TAV) encoded by SVBV, and can efficiently trans-activate the translation of downstream gfp mRNA in a bicistron derived from the SVBV. Furthermore, the P6 protein can trans-activate the expression of different bicistrons expressed by different caulimovirus promoters. The P6 protein encoded by SVBV from an infectious clone can also trans-activate the expression of bicistron. Through protein-protein interaction assays, we determined that the P6 protein could interact with the cell translation initiation factor FveIF3g of Fragaria vesca and co-localize with it in the nuclei of Nicotiana benthamiana cells. This interaction reduced the formation of P6 granules in cells and its trans-activation activity on translation.


Strawberry vein banding virus (SVBV) is an important plant virus infecting strawberries. SVBV
is now widely distributed in many countries worldwide, including Australia, Japan, multiple European countries, The United States of America, and China, and causes severe loss of strawberry production [1][2][3]. SVBV is known to be transmitted by several aphid species, including Chaetosiphon sp., in a semi-persistent manner [4]. Symptoms of SVBV infection in strawberry (Fragaria×ananassa cv. Sachinoka) plants are mainly retarded plant growth and fewer numbers of creeping stems [5]. The woodland strawberry (Fragaria vesca) is a diploid wild strawberry species that has been used as an indicator plant of SVBV infection assays. Symptoms of SVBV-infected F. vesca plants are mainly chlorosis along the leaf veins and lobular distortion [6]. When F. vesca plants are mix-infected with SVBV and Strawberry crinkle virus (SCV), Strawberry mottle virus (SMoV), Strawberry mild yellow edge virus (SMYEV), or Strawberry latent ring spot virus (SLRSV), the infected plants often show leaf distortion, chlorosis, plant stunting, and fruit yield reduction [7,8].
SVBV is a member of the genus Caulimovirus, family Caulimoviridae. The SVBV genome is a double-stranded DNA (dsDNA) with about 8000 nucleotides [9]. SVBV genomic DNA has seven open reading frames (ORFs) and shares similar genome structure with Cauliflower mosaic virus the GFP protein from the US71GFP bicistron, but the P6 could ( Figure S1A). Western blot analyses also revealed that the GFP protein was not detected in the pUS-US71GFP and pCAM-p19 or the pUS-US71GFP and pCAM-2b co-infiltrated leaves ( Figure S1B). These results indicate that the function of the RNA silencing suppressor does not promote the expression of the bicistron. Obviously, the P6 protein trans-activation activity is non-related with its VSR activity. The infiltrated leaves were photographed under the UV illumination at 3 days post agro-infiltration (dpai); (C) Western blot and Northern blot assays of the GFP protein and GFP mRNA accumulation in various infiltrated leaves. P1 through P6 indicate the six vectors expression SVBV P1, P2, P3, P4, P5 and P6 proteins. The same samples were used for these two assays. Western blot was probed with a GFP specific monoclonal antibody and the Northern blot was probed with a DIG-labeled GFP-specific probe. Coomassie blue staining was used to estimate protein loadings and ethidium bromide staining was used to estimate RNA loadings.

Plant Materials and Growth Conditions
N. benthamiana plants were grown inside an insect-free growth chamber set at 25 • C and a 16/8 h (light/dark) photoperiod. Plants at the 4-6 leaf stage were used for the experiments.

Agro-Infiltration and GFP Fluorescence Assay
Agrobacterium tumefaciens was transformed with the individual constructs described above and grown overnight in the YEP medium containing the appropriate antibiotics. A. tumefaciens cultures were pelleted and then resuspended individually in an infiltration buffer [10 mM MgCl 2 , 10 mM 2-(N-Morpholino) ethane sulfonic acid (MES), and 100 µM acetosyringone]. Individual A. tumefaciens cultures were adjusted to OD 600 = 1.2, or as indicated otherwise, and then mixed with the infiltration buffer by 1 to 1 volume for 3 h in dark conditions. The mixed A. tumefaciens suspension co-infiltrated into leaves of 7-8 leaf stage N. benthamiana plants using 1 mL needleless syringes. The agro-infiltrated leaves were examined for GFP expression under a hand-held 100 W, long-wave UV lamp (UV Products, Upland, CA, USA). The infiltrated leaves were photographed with a Canon EOS 700 D digital camera (Canon, Taiwan, China) with an Y48 yellow filter.

Confocal Microscopy
N. benthamiana leaves were infiltrated with A. tumefaciens cultures harboring specific combinations of constructs. At 70 h post agro-infiltration (hapi), a leaf disc (1.0 cm 2 ) was taken from an infiltrated leaf and examined under an OLYMPUS FluoView™ FV1000 confocal microscope (OLYMPUS, Tokyo, Japan). GFP fusion proteins were excited at 488 nm and the emitted signal was captured at 510 nm. In this study, 4,6-Diamidino-2-phenylindole (DAPI) (Sigma-Aldrich, St. Louis, MO, USA) was used to stain the nuclei in cells. All captured images were processed using the Adobe Photoshop software.

Protein Extraction and Western Blot Assays
The total protein was extracted from A. tumefaciens infiltrated leaf samples using the RIPA lysis buffer II (0.5g tissue/mL buffer) as instructed (Sangon Biotech, Shanghai, China). The extracts were centrifuged at 10,000 g for 5 min at 4 • C and the supernatant from each sample was collected. For Western blot assays, the total protein in each supernatant was separated in 10 % SDS-PAGE gels. After transferring protein bands to nitrocellulose membranes, the membranes were probed with a mouse anti-GFP monoclonal or a rabbit anti-P6 polyclonal antibody followed by an HRP-labeled goat anti-mouse IgG or anti-rabbit IgG antibody. Detection signal was visualized using the EasySee Western Blot Kit (TransGen Biotech, Beijing, China).

RNA Extraction and Northern Blot Assays
The total RNA was extracted from agro-infiltrated leaf samples using the Omini Plant RNA Plant Kit (CWBIO, Beijing, China). The synthesis of cDNA was conducted using 1.0 µg total RNA per 20 µL reaction using the PrimeScript™ RT reagent Kit with a gDNA Eraser (TaKaRa, Tokyo, Japan). For Northern blot analyses, the total RNA was first separated in 1.5% denatured agarose gels and then transferred onto Hybond-N+ membranes (GE Healthcare, Chicago, IL, USA). The membranes were hybridized with a DIG-labeled GFP specific DNA probe, and the labeling signal was detected using a labeling and detection starter Kit II (Roche, Basel, Switzerland).

Yeast Two-Hybrid and Bimolecular Fluorescence Complementation (BIFC) Assays
Interactions between P6 and FveIF3g, P6 and FvL18 or P6 and FvL24 were analyzed through Yeast two-hybrid (Y2H) assay as instructed by the manufacturer (Clontech, Mountain View, CA, USA). The yeast strain Y2H Gold was co-transformed with the pGAD-and pGBK-vectors and grown on an SD-Leu-Trp selective dropout medium for about 48 h in 28 • C. The selected positive colonies were transferred onto the SD/-Ade/-His/-Leu/-Trp plates and incubated at 28 • C for 72 h. For bimolecular fluorescence complementation (BiFC) assays, N. benthamiana leaves were co-infiltrated with A. tumefaciens cultures harboring YFP N -and YFP C -vectors using needless syringes. The agro-infiltrated N. benthamiana leaves were examined and photographed under the confocal microscope as described above.

SVBV P6 is a Trans-Activator of RNA Translation
To explore the mechanism controlling SVBV 35S RNA translation, bicistron expression vectors carrying an SVBV-US 35S promoter (pUS-US71GFP) were constructed. The full length ORF VII of SVBV-US served as the first cistron, and its ORF I 5 terminal 51 nts fused with a gfp gene served as the second cistron ( Figure 1A).
To determine which SVBV-encoded protein was a translation trans-activator, we cloned individual SVBV-US ORFs into the pCAM-2300 vector to produce pCAM-USP1, pCAM-USP2, pCAM-USP3, pCAM-USP4, pCAM-USP5, and pCAM-USP6. These vectors were individually co-infiltrated with pUS-US71GFP into N. benthamiana leaves. By 3 days post agro-infiltration (dpai), only the leaves co-infiltrated with pCAM-USP6 and pUS-US71GFP showed strong GFP fluorescence ( Figure 1B). No strong GFP fluorescence was observed in the leaves co-infiltrated with other constructs. Western blot assays showed a high amount of the GFP protein had accumulated in the leaves co-infiltrated with pCAM-USP6 and pUS-US71GFP. In contrast, only low levels of the GFP protein had accumulated in the leaves co-infiltrated with other constructs ( Figure 1C upper panel). Northern blot assays showed that GFP mRNA accumulated at similar levels in all co-infiltrated leaves ( Figure 1C).
A previous report showed that the SVBV P6 protein could suppress gfp gene silencing triggered by GFP ssRNA [11]. Tombusvirus p19 and CMV 2b are well known silencing suppressors. To exclude the effects of the suppressor functions from the translation, we tested whether the P6 protein could affect the expression of bicistron vectors. When pCAM-p19 or pCAM-2b were co-infiltrated with pUS-US71GFP into N. benthamiana leaves, no strong GFP fluorescence was observed by 3 dpai. In contrast, a strong GFP fluorescence was observed in the pCAM-USP6 and US71GFP co-expressed leaves. This indicated that neither the p19 nor the 2b proteins could trans-activate the expression of the GFP protein from the US71GFP bicistron, but the P6 could ( Figure S1A). Western blot analyses also revealed that the GFP protein was not detected in the pUS-US71GFP and pCAM-p19 or the pUS-US71GFP and pCAM-2b co-infiltrated leaves ( Figure S1B). These results indicate that the function of the RNA silencing suppressor does not promote the expression of the bicistron. Obviously, the P6 protein trans-activation activity is non-related with its VSR activity.

P6 Can Trans-Activate the Translation of the gfp RNA from Different Bicistrons
To further investigate whether the P6 protein can trans-activate the translation of different bicistrons, the CaMV 35S promoter (p35S-Ca71GFP) and the SVBV-SY 35S promoter (pSY-SY71GFP) were constructed. In these vectors, the full length ORF VII of CaMV or SVBV-SY served as the first cistron and their ORF I 5 terminal 51 nts fused with a gfp gene served as the second cistron ( Figure 2A). We then used the expression vectors from pUS-US71GFP, p35S-Ca71GFP, or pSY-SY71GFP bicistrons to mimic the expression of proteins from SVBV 35S polycistron RNA through the re-initiation of the translation of downstream cistrons. pCAM-USP6 was co-infiltrated with these three bicistronic constructs into N. benthamiana leaves. Leaves co-infiltrated with pCAM-USP6 and an empty vector (pCAM-2300) were used as a negative control. All of the leaves co-infiltrated with pUS-US71GFP and pCAM-USP6, p35S-Ca71GFP and pCAM-USP6, or pSY-SY71GFP and pCAM-USP6 showed strong GFP fluorescence under the UV illumination by 3 dpai. The leaves co-infiltrated with pCAM-2300 and pUS-US71GFP, pCAM-2300 and p35S-Ca71GFP, or pCAM-2300 and pSY-SY71GFP did not show GFP fluorescence ( Figure 2B,C). Western blot assays showed that high levels of the P6 and the GFP proteins had accumulated in the leaves co-infiltrated with pCAM-USP6 and pUS-US71GFP, p35S-Ca71GFP, or pSY-SY71GFP. Very low levels of the P6 and the GFP proteins had accumulated in the leaves co-infiltrated with pCAM-2300 and pUS-US71GFP, p35S-Ca71GFP, or pSY-SY71GFP ( Figure 2D, Upper panel). Northern blot assays showed that the gfp mRNA had accumulated to a similar level in various infiltrated leaves ( Figure 2D, Bottom panel). Therefore, we considered that Viruses 2018, 10, 717 7 of 18 P6 could trans-activate the translation of bicistronic vectors mimicking different-but functionally related-viruses, where the promoter and the first cistron originated from the same virus.

P6 Can Trans-Activate the Translation of gfp mRNA in Bicistrons Expressed Through a Heterozygous Promoter
To further demonstrate that the P6 protein can trans-activate the translation of gfp mRNA, we used bicistrons expressed through heterozygous promoters (i.e., p35S-SY71GFP, pUS-Ca71GFP and pUS-SY71GFP) ( Figure 3A). The three constructs were individually co-infiltrated with pCAM-USP6 into N. benthamiana leaves. Leaves infiltrated with pUS-US71GFP alone were used as negative controls. The results showed that strong GFP fluorescence was observed in the leaves co-infiltrated with pUS-US71GFP and pCAM-USP6, pUS-SY71GFP and pCAM-USP6, p35S-SY71GFP and pCAM-USP6, or pUS-Ca71GFP and pCAM-USP6 at 70 hpai. In this study, no GFP fluorescence was observed in the leaves infiltrated with pUS-US71GFP alone. Also, among various treatments, the leaves co-infiltrated with pUS-US71GFP and pCAM-USP6 showed the strongest green fluorescence ( Figure 3B). These results indicated that SVBV P6 could trans-activate the translation of bicistrons expressed through heterozygous promoters but the trans-activated translation activity was less efficient. empty vector (pCAM-2300) were used as a negative control. All of the leaves co-infiltrated with pUS-US71GFP and pCAM-USP6, p35S-Ca71GFP and pCAM-USP6, or pSY-SY71GFP and pCAM-USP6 showed strong GFP fluorescence under the UV illumination by 3 dpai. The leaves co-infiltrated with pCAM-2300 and pUS-US71GFP, pCAM-2300 and p35S-Ca71GFP, or pCAM-2300 and pSY-SY71GFP did not show GFP fluorescence ( Figure 2B and 2C). Western blot assays showed that high levels of the P6 and the GFP proteins had accumulated in the leaves co-infiltrated with pCAM-USP6 and pUS-US71GFP, p35S-Ca71GFP, or pSY-SY71GFP. Very low levels of the P6 and the GFP proteins had accumulated in the leaves co-infiltrated with pCAM-2300 and pUS-US71GFP, p35S-Ca71GFP, or pSY-SY71GFP ( Figure 2D, Upper panel). Northern blot assays showed that the gfp mRNA had accumulated to a similar level in various infiltrated leaves ( Figure 2D, Bottom panel). Therefore, we considered that P6 could trans-activate the translation of bicistronic vectors mimicking differentbut functionally related-viruses, where the promoter and the first cistron originated from the same virus. The Western blots were probed with a GFP or SVBV P6 specific antibody. The Northern blot was analyzed as described in Figure 1D).

P6 Can Trans-Activate the Translation of gfp mRNA in Bicistrons Expressed Through a Heterozygous Promoter
To further demonstrate that the P6 protein can trans-activate the translation of gfp mRNA, we used bicistrons expressed through heterozygous promoters (i.e., p35S-SY71GFP, pUS-Ca71GFP and pUS-SY71GFP) ( Figure 3A). The three constructs were individually co-infiltrated with pCAM-USP6 The Western blots were probed with a GFP or SVBV P6 specific antibody. The Northern blot was analyzed as described in Figure 1D).
with pUS-US71GFP and pCAM-USP6, pUS-SY71GFP and pCAM-USP6, p35S-SY71GFP and pCAM-USP6, or pUS-Ca71GFP and pCAM-USP6 at 70 hpai. In this study, no GFP fluorescence was observed in the leaves infiltrated with pUS-US71GFP alone. Also, among various treatments, the leaves coinfiltrated with pUS-US71GFP and pCAM-USP6 showed the strongest green fluorescence ( Figure  3B). These results indicated that SVBV P6 could trans-activate the translation of bicistrons expressed through heterozygous promoters but the trans-activated translation activity was less efficient.  3.4. P6 Protein Encoded by SVBV Can Also Trans-Activate the Translation of gfp mRNA F. vesca is the original host of the SVBV. Our previous study showed that agro-infiltration of the pSVBV-US infectious clone led to systemic infection, while as N. benthamiana is not its host, agro-infiltration with the same clone resulted in local viral protein accumulation only when it was co-infiltrated with a strong suppressor of RNA silencing (CMV 2b) [27]. In contrast, N. benthamiana plants agro-infiltrated with the same infectious clone did not show local and systemic infection. Interestingly, when the pSVBV-US infectious clone was co-infiltrated with a vector expressing a CMV 2b protein into N. benthamiana leaves, SVBV did accumulate in the infiltrated leaves. Western blot assays using an SVBV P6 specific antibody showed that the P6 protein was present in the infiltrated leaves. To determine if the SVBV-encoded P6 protein could also trans-activate the expression of the gfp mRNA in the US71GFP bicistron, we co-infiltrated three constructs (i.e., pUS-US71GFP, pSVBV-US, and pCAM-2b) into N. benthamiana leaves. Leaves co-infiltrated with pUS-US71GFP, empty vector pBINPLUS, and pCAM-2b were used as negative controls. Results showed that strong GFP fluorescence was observed in the leaves co-infiltrated with pUS-US71GFP, pSVBV-US, and pCAM-2b by 3 dpai, but not in the negative control leaves ( Figure 4A). Western blot assays confirmed that both P6 and GFP proteins had accumulated in the leaves co-infiltrated with pUS-US71GFP, pSVBV-US, and pCAM-2b. Very low amounts of GFP were detected in the control leaves ( Figure 4B). Northern blot assays showed that the accumulation of gfp RNA in the pUS-US71GFP, pSVBV-US, and pCAM-2b co-infiltrated leaves was similar to that in the negative control leaves ( Figure 4B). These results indicated that the translation of US71GFP bicistron could be trans-activated by the SVBV-encoded P6 protein.   The infiltrated plants were photographed under the UV illumination at 3 dpai; (B) Western blot assay of GFP and P6 protein accumulation, and Northern blot assay of GFP RNA accumulation in the infiltrated leaves. Lane 1, a sample from the pUS-US71GFP, pBinPLUS, and 2b co-infiltrated leaves. Lane 2, a sample from the pUS-US71GFP, pSVBV-US, and 2b co-infiltrated leaves. Lane 3, a sample from the pUS-US71GFP and pCAM2300 co-infiltrated leaves. Procedures of the Western blot and Northern blot assays were the same as described in Figure 1.

P6 Protein Can Interact with FveIF3g of F. vesca Both In Vitro and In Vivo
According to previous studies [25], we hypothesized that strawberry polyribosome subunit proteins L18, L24, and initiation factor eIF3g may have roles in the trans-activated translation of SVBV 35S RNA by the P6 protein. F. vesca FvL18, FvL24, and FveIF3g genes were individually cloned into Y2H vectors and then tested for the possible interactions between SVBV P6 and one of the three F. vesca proteins through Y2H assays. Results showed that the P6 protein did interact with FveIF3g but not with FvL18 or FvL24 ( Figure 5A). Interaction between the P6 and FveIF3g proteins was also confirmed by BiFC in N. benthamiana leaves ( Figure 5B). When N. benthamiana leaves were co-infiltrated with vectors expressing FveIF3g-mcherry or P6-GFP, red fluorescence from FveIF3g-mcherry and green fluorescence from P6-GFP were found to be co-localized in the infiltrated cells at 70 hpai ( Figure 5C), suggesting that FveIF3g could interact with the SVBV P6 protein in vivo.

FveIF3g Affects P6 Protein Accumulation and Inhibits the Trans-Activation Activity of P6 Protein
To test whether FveIF3g could interfere with P6 protein expression, we co-infiltrated N. benthamiana leaves with pCAM-P6:GFP and pCAM-FveIF3g or pCAM2300 (control). By 3 dpai, GFP fluorescence was observed in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g. However, the strength of GFP fluorescence was strongly reduced compared with that in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-2300. Confocal microscopy also showed that fewer and smaller P6:GFP granules were present in the cells co-infiltrated with pCAM-P6:GFP and pCAM-veIF3g than in the cells co-infiltrated with pCAM-P6:GFP and pCAM-2300 ( Figure 6A, Figure S2). Western blot analyses showed that much less P6 protein had accumulated in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g than had in the control leaves ( Figure 6B). Northern blot results also showed that less GFP RNA had accumulated in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g than had in the control leaves ( Figure 6C). Taken together, we conclude that FveIF3g can inhibit the accumulation of the P6:GFP fusion protein and the formation of granules in plant cells.

FveIF3g Affects P6 Protein Accumulation and Inhibits the Trans-Activation Activity of P6 Protein
To test whether FveIF3g could interfere with P6 protein expression, we co-infiltrated N. benthamiana leaves with pCAM-P6:GFP and pCAM-FveIF3g or pCAM2300 (control). By 3 dpai, GFP fluorescence was observed in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g. However, the strength of GFP fluorescence was strongly reduced compared with that in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-2300. Confocal microscopy also showed that fewer and smaller P6:GFP granules were present in the cells co-infiltrated with pCAM-P6:GFP and pCAM-veIF3g than in the cells co-infiltrated with pCAM-P6:GFP and pCAM-2300 ( Figure 6A, Figure S2). Western blot analyses showed that much less P6 protein had accumulated in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g than had in the control leaves ( Figure 6B). Northern blot results also showed that less GFP RNA had accumulated in the leaves co-infiltrated with pCAM-P6:GFP and pCAM-FveIF3g than had in the control leaves ( Figure 6C). Taken together, we conclude that FveIF3g can inhibit the accumulation of the P6:GFP fusion protein and the formation of granules in plant cells.
We then examined the effect of FveIF3g on P6 trans-activating activity by co-infiltrating N. benthamiana leaves with pUS-US71GFP, pCAM-USP6, and pCAM-FveIF3g or pUS-US71GFP, pCAM-USP6, and pCAM-2300 (control). This was followed by visual observations under the UV illumination at 3 dpai. GFP fluorescence in the leaves co-infiltrated with pUS-US71GFP, pCAM-USP6, and pCAM-FveIF3g was much weaker than it was in the control leaves. Western blot results showed again that the expression level of the P6 protein was much lower in the leaves co-infiltrated with pUS-US71GFP, pCAM-USP6, and pCAM-FveIF3g compared with the control leaves ( Figure 6D,E). Taken together, we conclude that the expression of FveIF3g can suppress the P6 protein accumulation, leading to a significant reduction of P6 translation trans-activation activity. USP6, and pCAM-FveIF3g was much weaker than it was in the control leaves. Western blot results showed again that the expression level of the P6 protein was much lower in the leaves co-infiltrated with pUS-US71GFP, pCAM-USP6, and pCAM-FveIF3g compared with the control leaves ( Figure 6D and 6E). Taken together, we conclude that the expression of FveIF3g can suppress the P6 protein accumulation, leading to a significant reduction of P6 translation trans-activation activity.

Discussion
CaMV 35S RNA has a 600 nt leader sequence containing a stem-loop structure in the middle and a short ORF upstream of the stem-loop structure [28]. During virus replication, the ribosome initiates RNA translation along the 35S RNA. After the ribosome slides through the termination codon of the first short ORF, the stem-loop structure will stop the ribosome in order to continue the translation. With the help of CaMV P6, however, the ribosome can scan-reinitiate and shunt to the 3′ end of the stem-loop and then continue to slide through the 35S RNA for downstream ORF translations [29]. The CaMV P6 is also known to have other important roles in virus replication and infection in plants [30].
In this study, we constructed a series of bicistron expression vectors to investigate the mechanism controlling the translation of SVBV 35S RNA polycistron. In these expression vectors, the (B) Northern blot assay of GFP RNA accumulation in the infiltrated leaves; (C) Western blot assay of P6 protein accumulation in the infiltrated leaves; (D) N. benthamiana leaves were co-infiltrated with pUS-US71GFP, pCAM-USP6 (P6), and p2300-FveIF3g (FveIF3g), or pUS-US71GFP, P6, and pCAM-2300 (Vec). The infiltrated plants were photographed under the UV light at 3 dpai. GFP fluorescence in the pUS-US71GFP, P6, and FveIF3g co-infiltrated leaves was weaker than that in the pUS-US71GFP, P6, and Vec co-infiltrated leaves; (E) Western blot assays of GFP and P6 protein accumulation in the co-infiltrated leaves at 3 dpai. Coomassie brilliant blue staining was used to estimate the sample loadings.

Discussion
CaMV 35S RNA has a 600 nt leader sequence containing a stem-loop structure in the middle and a short ORF upstream of the stem-loop structure [28]. During virus replication, the ribosome initiates RNA translation along the 35S RNA. After the ribosome slides through the termination codon of the first short ORF, the stem-loop structure will stop the ribosome in order to continue the translation. With the help of CaMV P6, however, the ribosome can scan-reinitiate and shunt to the 3 end of the stem-loop and then continue to slide through the 35S RNA for downstream ORF translations [29]. The CaMV P6 is also known to have other important roles in virus replication and infection in plants [30].
In this study, we constructed a series of bicistron expression vectors to investigate the mechanism controlling the translation of SVBV 35S RNA polycistron. In these expression vectors, the full length SVBV ORF VII was used to serve as the first cistron and the ORF I 5 terminal 51 nts fused with a gfp gene was used to serve as the second cistron. We reasoned that if the translation of the gfp gene in these bicistrons was activated in N. benthamiana leaf cells, GFP fluorescence should be observed. Our results showed that the leaves infiltrated with the bicistron expression vector alone did not produce green fluorescence and the GFP protein was not detected in the infiltrated leaves by Western blot assays. When the bicistron expression vector was co-expressed with individual SVBV proteins in N. benthamiana leaves, the leaves co-expressing SVBV P6 and the bicistron exhibited strong green fluorescence, and both SVBV P6 and GFP proteins were detected in the infiltrated leaves, indicating that SVBV P6 could trans-activate the translation of the bicistron. In contrast, none of the other SVBV-encoded proteins showed the trans-activation activity in this study.
A previous report indicated that the virus-encoded RNA silencing suppressor could suppress the immune response of a host plant against virus infection, and promote the expression of exogenous genes in plants [31]. Our previous report also showed that the SVBV P6 protein is a multifunctional protein that can act as a viral suppressor of RNA silencing (VSR) to suppress gfp gene silencing induced by a single-stranded gfp RNA [11]. In this study, we determined that the P6 protein could also trans-activate the expression of the GFP protein when it was co-infiltrated with the pUS-US71GFP bicistron vector into N. benthamiana leaves. It is still unclear whether the P6 trans-activation activity is associated with its VSR activity. To investigate this possibility, we compared SVBV P6 with other two VSRs-the TBSV P19 and the CMV 2b protein [32,33]. The results showed that neither the p19 nor the 2b could trans-activate the expression of the GFP protein. This finding indicates that the P6 protein trans-activation activity is independent from its VSR activity.
To further determine the trans-activation activity of SVBV P6, we constructed different bicistron expression vectors driven by different caulimovirus promoters (i.e., pUS-US71GFP, p35S-Ca71GFP, pSY-SY71GFP, p35S-SY71GFP, pUS-Ca71GFP, and pUS-SY71GFP). Agro-infiltration assays showed that the P6 protein was able to trans-activate the translation of all bicistrons, suggesting that the trans-activation ability of the P6 protein was conserved. We speculate that the translation mechanism controlling polycistron translation is the same among caulimoviruses. In this study, we also found that the trans-activation activity of the P6 protein on the bicistrons expressed through the SVBV promoter was stronger than that of the bicistrons expressed through the other promoters. We speculate that this difference is caused by a weaker binding of the SVBV P6 protein to other 35S promoters. Northern blot results confirmed that all the agro-infiltrated leaves accumulated similar amounts of GFP mRNA, indicating that the P6 protein acted at the translation step but not the transcription step.
The SVBV infectious clone, pSVBV-US, could cause local and systemic infection in F. vesca but not in N. benthamiana plants. In contrast, SVBV infection was observed in the N. benthamiana leaves co-infiltrated with pSVBV-US and a vector expressing CMV 2b. In this study, strong GFP fluorescence was observed in the leaves co-infiltrated with pUS-US71GFP, pSVBV-US and pCAM-2b. These results suggest that the translation of bicistron could also be trans-activated by the P6 protein produced during SVBV infection in N. benthamiana.
The 80S ribosome is comprised of 40S and 60S ribosomal subunits and can initiate the translation of CaMV 35S RNA during CaMV infection in plants [14]. After finishing the translation of the previous ORF, the 40S ribosomal subunit can maintain the state of translation on the 35S RNA. The CaMV P6 protein reportedly binds the eIF3g and 40S ribosomal subunits, and then recruits the 60S ribosomal subunit to continue the translation of downstream OFRs [34]. Therefore, the interaction between the P6 and eIF3g proteins is needed to start the translation of the later cistrons. In this study, we also investigated the roles of FvL18, FvL24, and FveIF3g during the translation trans-activation by the SVBV P6 protein. Our results showed that only FveIF3g could interact with the P6 and co-localized with the P6 in the nuclei. It is possible that FveIF3g can interfere with ORF VI expression in the nuclei, leading to the reduction of the P6 protein accumulation and the attenuation of the trans-activation of downstream gene translation. The molecular mechanism(s) that cause FveIF3g to inhibit the accumulation of the P6 protein in cells requires future research.
Each plant virus genome encodes at least three proteins. Therefore, all the genomic RNAs of single component RNA viruses are polycistrons. The cistrons located at the 5 end can be directly translated, but the translation of the cistrons that are far away from the 5 end need other expression strategies. Subgenomic RNA (Tobamovirus) and polyprotein cleavage (Potyvirus) are the most common strategies for genomic expression in RNA viruses. Caulimoviruses are a kind of dsDNA virus. The polycistron RNAs derived from dsDNA express proteins through trans-activation, which is totally different from subgenomic RNA and polyprotein cleavage. The P6 protein encoded by CaMV ORF VI has been proven to be a trans-activator, and it can efficiently trans-activate the translation of downstream mRNA in a bicistron. Our study showed that the P6 protein encoded by SVBV ORF VI had similar trans-activation function. Interestingly, CaMV ORF VI and SVBV ORF VI share a very low nucleotide sequence similarity-about 40%. However, the SVBV P6 protein not only trans-activated the translation of the SVBV bicistron, but the classical CaMV bicistron as well. This implies that the regions associated with trans-activation in CaMV and SVBV might be very short, possibly controlled by only one or several motifs. The above results indicate that trans-activation might be a common strategy for the genomic expression of caulimoviruses. Our research enriches the understanding of genomic expression strategies of plant DNA viruses.
Funding: This research was financially supported by the Grants from the National Natural Science Fund of China (31671999, 31801700).

Conflicts of Interest:
The authors declare no conflict of interest.