The conserved aromatic residue W¹²² is a determinant of potyviral coat protein stability, replication, and cell‐to‐cell movement in plants

Abstract

Coat proteins (CPs) play critical roles in potyvirus cell‐to‐cell movement. However, the underlying mechanism controlling them remains unclear. Here, we show that substitutions of alanine, glutamic acid, or lysine for the conserved residue tryptophan at position 122 (W¹²²) in tobacco vein banding mosaic virus (TVBMV) CP abolished virus cell‐to‐cell movement in Nicotiana benthamiana plants. In agroinfiltrated N. benthamiana leaf patches, both the CP and RNA accumulation levels of three W¹²² mutant viruses were significantly reduced compared with those of wild‐type TVBMV, and CP accumulated to a low level similar to that of a replication‐deficient mutant. The results of polyprotein transient expression experiments indicated that CP instability was responsible for the significantly low CP accumulation levels of the three W¹²² mutant viruses. The substitution of W¹²² did not affect CP plasmodesmata localization or virus particle formation; however, the substitution significantly reduced the number of virus particles. The wild‐type TVBMV CP could complement the reduced replication and abolished cell‐to‐cell movement of the mutant viruses. When the codon for W¹²² was mutated to that for a different aromatic residue, phenylalanine or tyrosine, the resultant mutant viruses moved systemically and accumulated up to 80% of the wild‐type TVBMV level. Similar results were obtained for the corresponding amino acids of W¹²² in the watermelon mosaic virus and potato virus Y CPs. Therefore, we conclude that the aromatic ring in W¹²² in the core domain of the potyviral CP is critical for cell‐to‐cell movement through the effects on CP stability and viral replication.

The aromatic ring of W¹²² and corresponding residues is crucial for the coat protein stability, replication, and cell‐to‐cell movement of potyviruses.

1. INTRODUCTION

The genus Potyvirus, belonging to the family Potyviridae, is the largest genus of plant RNA viruses, consisting of more than 160 species (Wylie et al., 2018). Many potyviruses cause severe economic losses in crop production (García et al., 2014; Scholthof et al., 2011; Valkonen, 2007). The potyvirus genome is a positive‐sense single‐stranded RNA molecule with 9,400 to 11,000 nucleotides. The potyviral genomic RNA has a polyadenylated tail at its 3′ end and covalently links with a genome‐linked protein (VPg) at its 5′ end (Tavert‐Roudet et al., 2017; Wylie et al., 2018). The potyvirus genome encodes two polyproteins that are processed into 11 mature proteins by three self‐encoded proteinases (Chung et al., 2008; Olspert et al., 2015; Revers & García, 2015; Rodamilans et al., 2015).

Cell‐to‐cell movement is a critical step for plant viruses to establish systemic infection (Ritzenthaler, 2011; Schoelz et al., 2011). To move between cells, plant viruses take advantage of plasmodesmata (PD), which are microscopic channels for plant communication between adjacent cells. Potyviruses do not encode a specific movement protein. However, the second 6 kDa protein (6K2) (Grangeon et al., 2013; Jiang et al., 2015), the third protein (P3) (Chai et al., 2020), the frame‐shift protein P3N‐PIPO produced by transcriptional slippage (Cheng et al., 2017; Geng et al., 2015; Olspert et al., 2015; Rodamilans et al., 2015; Wei et al., 2010), the cylindrical inclusion protein (CI) (Carrington et al., 1998; Deng et al., 2015; Movahed et al., 2017; Roberts et al., 1998; Rodríguez‐Cerezo et al., 1997), and the coat protein (CP) (Dolja et al., 1994, 1995) are known to be involved in potyviral cell‐to‐cell movement.

Potyviral CPs contain three domains: the N‐terminal domain with 17–78 residues (the numbers vary among viruses), the core domain with 213–218 residues (starting with the KDK/D residues at the N‐terminus and ending with the TER/H residues at the C‐terminus) (Shukla et al., 1988), and the C‐terminal domain with 17–21 residues. Deletion of the N‐ or C‐ terminal domains of the tobacco etch virus (TEV) CP reduces its cell‐to‐cell movement in plants (Dolja et al., 1994, 1995). Alteration in the net charge of the N‐terminal domain of the zucchini yellow mosaic virus CP, or mutation of the charged arginine residue at position 245 (R²⁴⁵), histidine at position 246 (H²⁴⁶), or aspartic acid at position 250 (D²⁵⁰), which are at the border between the core and the C‐terminal domains of the soybean mosaic virus CP, nearly abolishes viral cell‐to‐cell movement in plants (Kimalov, 2004; Seo et al., 2013). The phosphorylation sites in the T²⁴²TSEED²⁴⁷ motif in CP are critical for the replication of potato virus A (Ivanov et al., 2003; Lõhmus et al., 2017). Mutation of the residues in serine (S¹²²), R¹⁵⁴, or D¹⁹⁸ in the TEV CP completely abolishes TEV particle formation and cell‐to‐cell movement, suggesting that TEV moves between cells in the form of viral particles (Dolja et al., 1994, 1995). Although the core domain of the potyvirus CP is known to be involved in the viral cell‐to‐cell movement, the roles of the conserved amino acid residues in this domain in viral cell‐to‐cell movement remain largely unknown.

In this study, we determined that the aromatic ring of the conserved tryptophan residue at position 122 (W¹²²) in the potyvirus CP core domain plays an essential role in viral cell‐to‐cell movement through maintaining viral replication and CP stability.

2. RESULTS

2.1. Mutations of W¹²² abolish TVBMV cell‐to‐cell and systemic movement in Nicotiana benthamiana plants

To investigate the effects of conserved amino acid residues in the TVBMV CP on virus infection, we first aligned the CP sequences of 139 potyviruses and identified the completely conserved aromatic residue tryptophan (W) at position 122 (W¹²²; numbered according to the TVBMV CP sequence) (Figure S1). We then mutagenized the codon for W¹²² in the pCamTVBMV‐GFP, an TVBMV infectious clone carrying the green fluorescent protein (GFP) gene, to the codons for residue alanine (A), glutamic acid (E), or lysine (K) to produce the plasmids pCamTVBMV^W122A‐GFP, pCamTVBMV^W122E‐GFP, and pCamTVBMV^W122K‐GFP, respectively (Figure 1a). Agrobacterium cultures carrying pCamTVBMV‐GFP or one of the three mutant plasmids were individually infiltrated into two expanded leaves of each assayed N. benthamiana plant to produce the viruses TVBMV‐GFP, TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, and TVBMV^W122K‐GFP, respectively. At 7 days postagroinfiltration (dpai), the TVBMV‐GFP‐infected plants showed mosaic and epinasty symptoms in their systemically infected leaves, but the three mutant viruses did not (Figure 1b). When the infiltrated plants were examined under a UV light, the TVBMV‐GFP‐infected plants emitted a strong green fluorescence in their infiltrated and systemically infected leaves. This green fluorescence was only observed in the leaves infiltrated with pCamTVBMV^W122A‐GFP, pCamTVBMV^W122E‐GFP, and pCamTVBMV^W122K‐GFP (Figure 1b). The results of the reverse transcription (RT)‐PCR and western blot assays showed that TVBMV RNA and CP had accumulated in the systemically infected leaves of the TVBMV‐GFP‐inoculated plants, but not in the systemically infected leaves of the plants inoculated with one of the three mutant viruses (Figure 1c,d). To validate the importance of W¹²² in TVBMV systemic movement, we changed the codon of nonconserved residues arginine at position 62 (R⁶²) to that for A, cysteine (C), or E; serine at position 92 (S⁹²) to that for A, aspartic acid (D), or histidine (H); and K at position 246 (K²⁴⁶) to that for A, C, or E in pCamTVBMV‐GFP. These mutant plasmids were individually infiltrated into N. benthamiana leaves and these mutations did not inhibit viral systemic infection (Figure S2), indicating that W¹²² in the CP core domain was crucial for TVBMV systemic infection in N. benthamiana plants.

FIGURE 1.

Conserved residue W¹²² in tobacco vein banding mosaic virus coat protein (TVBMV CP) is crucial for viral cell‐to‐cell and systemic movement. (a) A schematic diagram showing the genome organization of TVBMV in pCamTVBMV‐GFP. The N‐ and C‐terminal domains of TVBMV CP are in grey, and the core domain is in black. The relative position of W¹²² is indicated with a red arrowhead. The site‐directed mutagenized plasmids (left column) and viruses (right column) are shown in the red‐lined box. pCamTVBMV‐GFP, pCamTVBMV^W122A‐GFP, pCamTVBMV^W122E‐GFP, and pCamTVBMV^W122K‐GFP plasmids were individually agroinfiltrated into Nicotiana benthamiana plants. (b) Symptoms (upper panel) and green fluorescent protein (GFP) green fluorescence (lower panel) in various assayed N. benthamiana plants at 7 days postagroinfiltration (dpai) are shown. (c) and (d) Accumulation levels of TVBMV RNA and CP in the systemic leaves of the assayed plants were determined at 7 dpai through reverse transcription‐PCR and western blot assay, respectively. The expression of the Nbef1a gene was used as an internal control during RT‐PCR. Ab: antibody. The Coomassie brilliant blue‐stained RuBisCO large subunit protein (Rubi) was used to show sample loadings. (e) and (f) Cell‐to‐cell movement of TVBMV‐GFP and its three W¹²² mutants in N. benthamiana leaves at 132 hr postagroinfiltration (hpai). The pictures were taken under a confocal microscope at 132 hpai. Because each TVBMV‐GFP infection focus contained too many cells with GFP fluorescence, its number is not shown in (f). The values are presented as means ± SD from 30 infection foci per treatment. Statistical significance between treatments was determined using Duncan's multiple range test (p < .05)

Next, we investigated the role of W¹²² in TVBMV cell‐to‐cell movement. Agrobacterium cultures carrying pCamTVBMV‐GFP or one of its three W¹²² mutant plasmids were diluted to OD₆₀₀ = 0.0001 and individually infiltrated into N. benthamiana leaves. GFP fluorescence from TVBMV‐GFP was observed in clusters of multiple cells at 132 hr postagroinfiltration (hpai). However, the GFP fluorescence from TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, or TVBMV^W122K‐GFP was confined to single leaf cells (Figure 1e,f), indicating that W¹²² was also crucial for TVBMV cell‐to‐cell movement.

2.2. Mutations of W¹²² inhibit TVBMV replication

Cell‐to‐cell movement and replication are two important and interconnected steps during potyvirus infection (Chai et al., 2020; Cui & Wang, 2016). To determine whether W¹²² has a role in TVBMV replication, we infiltrated N. benthamiana leaves with pCamTVBMV‐GFP or one of the three W¹²² mutant plasmids. Plants infiltrated with pCamTVBMV^NIbΔGDD‐GFP (produce replication‐deficient mutant TVBMV^NIbΔGDD‐GFP; Geng et al., 2017) or pCamTVBMV^CPSTOP‐GFP (producing TVBMV^CPSTOP‐GFP that did not produce CP) were used as controls. The infiltrated leaves were harvested at 60 hpai and analysed for GFP or CP accumulation using enzyme‐linked immunosorbent assay (ELISA). The results showed that the accumulation levels of GFP in the TVBMV^W122A‐GFP‐, TVBMV^W122E‐GFP‐, or TVBMV^W122K‐GFP‐inoculated leaves were significantly reduced compared with that of the TVBMV‐GFP‐inoculated leaves. The accumulation levels of GFP in the TVBMV^NIbΔGDD‐GFP‐ or TVBMV^CPSTOP‐GFP‐inoculated leaves were the lowest (Figure 2a). The accumulation levels of CP in the TVBMV^W122A‐GFP‐, TVBMV^W122E‐GFP‐, and TVBMV^W122K‐GFP‐inoculated leaf tissues were also significantly reduced compared with that of the TVBMV‐GFP‐inoculated leaf tissues and reached a similar level to that of mutant TVBMV^NIbΔGDD‐GFP‐inoculated leaf tissues (Figure 2b). Similar results were obtained using the leaf samples harvested at 108 hpai (Figure S3).

FIGURE 2.

Conserved residue W¹²² in tobacco vein banding mosaic virus coat protein (TVBMV CP) is crucial for viral replication. The effect of W¹²² on TVBMV replication was tested by agroinfiltration of pCamTVBMV‐GFP, pCamTVBMV^W122A‐GFP, pCamTVBMV^W122E‐GFP, pCamTVBMV^W122K‐GFP, pCamTVBMV^NIbΔGDD‐GFP, and pCamTVBMV^CPSTOP‐GFP individually into Nicotiana benthamiana leaves. (a) and (b) The accumulation of green fluorescent protein (GFP) and TVBMV CP in various infiltrated leaf samples was determined through ELISA at 60 hr postagroinfiltration (hpai). The replication‐deficient TVBMV^NIbΔGDD‐GFP and TVBMV^CPSTOP‐GFP mutants were included as controls. (c) and (d) The accumulation of TVBMV (+)RNA and (−)RNA in various infiltrated leaf samples was determined through quantitative reverse transcription PCR using specific primers at 60 hpai. The expression of Nbef1a was used as an internal control. The results are presented as means ± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan's multiple range test (p < .05)

The accumulation levels of the TVBMV plus‐strand (+)RNA and minus‐strand (−)RNA in the infiltrated leaf tissues were determined using RT‐quantitative PCR (RT‐qPCR). The results showed that the accumulation levels of (+)RNA and (−)RNA in the TVBMV^W122A‐GFP‐, TVBMV^W122E‐GFP‐, or TVBMV^W122K‐GFP‐inoculated leaf tissues were significantly lower than that in the TVBMV‐GFP‐inoculated leaf tissues, but higher than those in the TVBMV^NIbΔGDD‐GFP‐ or the TVBMV^CPSTOP‐GFP‐inoculated leaf tissues (Figure 2c,d), indicating that W¹²² in the CP could also affect TVBMV replication.

2.3. W¹²² is crucial for TVBMV CP stability

To investigate why the three W¹²² mutants accumulated significantly less CP than TVBMV‐GFP, we constructed two vectors to transiently express polyprotein, NIaPro:HA‐NIb:GFP:CP and NIaPro:HA‐NIb:GFP:CP^W122A, in N. benthamiana leaves through agroinfiltration (Figure 3a). Each of the polyproteins could be self‐cleaved by NIaPro into four mature proteins: NIaPro, HA‐NIb, GFP, and CP (or the mutant CP^W122A). N. benthamiana leaves were infiltrated with mixed A grobacterium cultures harbouring pCamNIaPro:HA‐NIb:GFP:CP and pBinP19, or pCamNIaPro:HA‐NIb:GFP:CP^W122A and pBinP19. By 4 dpai, and the intensity of GFP fluorescence in the pCamNIaPro:HA‐NIb:GFP:CP^W122A and pBinP19 coinfiltrated N. benthamiana leaf patches was similar to that in the pCamNIaPro:HA‐NIb:GFP:CP and pBinP19 coinfiltrated leaf patches (Figure 3b). Western blot results showed that the accumulation levels of HA‐NIb and GFP in the pCamNIaPro:HA‐NIb:GFP:CP^W122A and pBinP19 coinfiltrated leaf patches were similar to those in the pCamNIaPro:HA‐NIb:GFP:CP and pBinP19 coinfiltrated leaf patches (Figure 3c). However, the accumulation level of CP in the pCamNIaPro:HA‐NIb:GFP:CP^W122A and pBinP19 coinfiltrated leaf patches was only one‐ninth of that in the pCamNIaPro:HA‐NIb:GFP:CP and pBinP19 coinfiltrated leaf patches (Figure 3c). These results suggest that mutation of W¹²² affects CP stability.

FIGURE 3.

Effect of W¹²² on tobacco vein banding mosaic virus coat protein (TVBMV CP) stability. (a) A schematic diagram showing the NIaPro:HA‐NIb:GFP:CP fragment in the pCamNIaPro:HA‐NIb:GFP:CP vector. HA‐NIb, GFP, and CP are cleaved from the polyprotein by NIaPro. (b) Nicotiana benthamiana leaves agroinfiltrated with pCamNIaPro:HA‐NIb:GFP:CP and pBinP19 or pCamNIaPro:HA‐NIb:GFP:CP^W122A and pBinP19 were photographed under a UV light at 4 days postagroinfiltration (dpai). (c) The accumulation of HA‐NIb, green fluorescent protein (GFP), and CP in the infiltrated leaf patches was determined through western blot assay using a HA (upper panel), GFP (middle panel), or TVBMV CP (lower panel) specific antibody at 4 dpai. Ab, antibody. The Coomassie brilliant blue‐stained RuBisCO large subunit protein (Rubi) was used to show sample loadings. (d) A schematic diagram showing plasmids expressing a GFP‐TVBMVCP or GFP‐TVBMVCP^W122A mutant fusion. (e) A time‐course study of GFP‐TVBMVCP and GFP‐TVBMVCP^W122A expression in the infiltrated leaf patches at 30, 36, 48, 60, 72, 84, 96, and 120 hpai. (f) Analysis of GFP intensity in the pCamGFP‐TVBMVCP and pCamGFP‐TVBMVCP^W122A‐infiltrated leaf patches at different time points using a microplate reader set at 485/10 nm (excitation wavelength) and 535/10 nm (emission wavelength). The values are presented as means ± SD from three biological replicates per treatment

To further investigate the stability of the W¹²² mutant CP, we constructed pCamGFP‐TVBMVCP and pCamGFP‐TVBMVCP^W122A to express a GFP‐TVBMVCP fusion and a GFP‐TVBMVCP^W122A fusion, respectively (Figure 3d). N. benthamiana leaves were infiltrated with mixed A grobacterium cultures harbouring pCamGFP‐TVBMVCP and pBinP19, or pCamGFP‐TVBMVCP^W122A and pBinP19. The infiltrated leaves were photographed and analysed at different times. The results showed that the intensity of GFP fluorescence in the leaf patches coexpressing GFP‐TVBMVCP and P19 continued to increase from 30 to 120 hpai, whereas the intensity of GFP fluorescence in the leaf patches coexpressing GFP‐TVBMVCP^W122A and P19 remained low (Figure 3e,f). We also constructed plasmids pCamTVBMVCP^WT and pCamTVBMVCP^W122A to transiently express TVBMVCP^WT and TVBMVCPW^122A in N. benthamiana leaves. Western blot results showed that, at 4 dpai, the accumulation levels of TVBMVCP^WT and TVBMVCP^W122A in the assayed leaf patches agreed with the results for GFP shown in Figure 3e,f (Figure S4). The expression of His‐TVBMVCP and His‐TVBMVCP^W122A in Escherichia coli cells also indicated that the mutation of W¹²² to A affected the stability of TVBMV CP (Figure S5). All these results indicate that mutation of W¹²² affects CP stability.

2.4. The mutation of W¹²² does not affect TVBMV particle formation

Previous studies have shown that virus particle formation is vital during potyviral cell‐to‐cell movement (Dolja et al., 1994). To determine whether the mutation of W¹²² could affect virus particle formation, we expressed TVBMV‐GFP, TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, and TVBMV^W122K‐GFP in N. benthamiana leaves. At 5 dpai, the infiltrated leaves were collected and used for virus particle purification. When the purified samples were negatively stained with 2% uranyl acetate and examined under a transmission electron microscope, virus particles similar to that of TVBMV‐GFP were observed in the TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, and TVBMV^W122K‐GFP purified samples, indicating that the mutation of W¹²² did not change the ability of CP to encapsidate viral RNA (Figure 4a). However, the numbers of virus particles in the TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, and TVBMV^W122K‐GFP purification samples were significantly reduced compared with that in the TVBMV‐GFP purification sample (Figure 4b).

FIGURE 4.

Effect of W¹²² on TVBMV particle formation. (a) Particles of TVBMV‐GFP, TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, and TVBMV^W122K‐GFP. The virus particles were purified from the infiltrated leaf tissues, negatively stained with 2% uranyl acetate, and then examined under a transmission electron microscope. (b) The total numbers of virus particles in fields of 70 μm² were counted for each treatment and are presented as the means of virus particles per field ± SD from six fields per treatment

2.5. The mutation of W¹²² does not affect the ability of CP to target PD

Potyviral CP, P3N‐PIPO, and CI are known to target PD in cell walls and are necessary for the potyviral cell‐to‐cell movement (Wei et al., 2010). To investigate the effects of W¹²² on TVBMV CP subcellular localization, we coinfiltrated N. benthamiana leaves with mixed A grobacterium cultures harbouring combinations of four constructs, including pCamTVBMV, pCamP3N‐PIPO, pCamCI‐DsRed, and pCamGFP‐TVBMVCP or pCamGFP‐TVBMVCP^W122A. At 48 hpai, analyses of infiltrated leaf patches under a confocal microscope showed that GFP‐TVBMVCP and GFP‐TVBMVCP^W122A could colocalize with CI‐DsRed in the cell periphery targeted to PD (Figure 5). Our previous results showed that TVBMV CI can localize at PD in the presence of TVBMV P3N‐PIPO (Geng et al., 2015). These results suggest that GFP‐TVBMVCP and GFP‐TVBMVCP^W122A could target PD with CI‐DsRed and P3N‐PIPO during TVBMV infection.

FIGURE 5.

Effect of W¹²² on CP subcellular localization. In this experiment, we coinfiltrated Nicotiana benthamiana leaves with mixed A grobacterium cultures harbouring various combinations of constructs: (a) pCamGFP‐TVBMVCP, pCamTVBMV, pCamP3N‐PIPO, and pCamCI‐DsRed; (b) pCamGFP‐TVBMVCP^W122A, pCamTVBMV, pCamP3N‐PIPO, and pCamCI‐DsRed. The infiltrated leaf tissues were collected at 48 hr postagroinfiltration (hpai) and examined under a confocal microscope. Both GFP‐TVBMVCP and GFP‐TVBMVCP^W122A were colocalized with CI‐DsRed at the periphery of N. benthamiana leaf cells in the presence of TVBMV P3N‐PIPO and TVBMV infection. White arrows indicate the points with both GFP‐TVBMVCP and CI‐DsRed or GFP‐TVBMVCP^W122A and CI‐DsRed. Bars = 5 μm. Pictures were photographed at 60 hpai under a confocal microscope

2.6. Wild‐type CP, but not the W122A mutant CP, rescue TVBMV^W122A‐GFP replication and cell‐to‐cell movement

To determine whether TVBMVCP^WT (wild‐type) or TVBMVCP^W122A could rescue the defective cell‐to‐cell movement of TVBMV^W122A‐GFP, we infiltrated N. benthamiana leaves with a mixture of three A grobacterium cultures harbouring pCamTVBMV^W122A‐GFP, pCamTVBMVCP^WT, or pCamTVBMVCP^W122A, and pBinP19. The results showed that, by 132 hpai, N. benthamiana leaves coexpressing TVBMV^W122A‐GFP, TVBMVCP^WT, and P19 showed GFP fluorescence in multiple cells, indicating that TVBMVCP^WT rescued the cell‐to‐cell movement of TVBMV^W122A‐GFP (Figure 6a,b). In contrast, the N. benthamiana leaves coexpressing TVBMV^W122A‐GFP, TVBMVCP^W122A, and P19 showed GFP fluorescence in single cells only, indicating that TVBMVCP^W122A could not rescue the cell‐to‐cell movement of TVBMV^W122A‐GFP.

FIGURE 6.

Wild‐type (WT) CP could rescue the defective cell‐to‐cell movement and reduced replication of TVBMV^W122‐GFP in trans. (a) Analysis of trans‐complemented cell‐to‐cell movement by a movement‐defective TVBMV^W122A‐GFP. Nicotiana benthamiana leaves were coinfiltrated with a mixture of three A grobacterium cultures carrying pCamTVBMV^W122A‐GFP or pCamTVBMV^CPSTOP‐GFP, pCamTVBMVCP^WT or pCamTVBMVCP^W122A, and pBinP19. The OD₆₀₀ values were adjusted to 0.0003 for cultures carrying plasmids pCamTVBMV^W122A‐GFP or pCamTVBMV^CPSTOP‐GFP, and 0.6 for other plasmids before mixing. The infiltrated leaf tissues were harvested at 132 hr postagroinfiltration (hpai) and examined under a confocal microscope. Movement‐ and replication‐defective mutant TVBMV^CPSTOP‐GFP was included in the experiment as a control. Bars = 50 μm. (b) The numbers of cells with GFP fluorescence per infection foci per treatment at 132 hpai. The values are means ± SD from 30 infection foci per treatment. (c) The relative accumulation levels of TVBMV (−)RNA in the agroinfiltrated N. benthamiana leaf tissues at 60 hpai. N. benthamiana leaves were infiltrated with a mixture of three A grobacterium cultures carrying pCamTVBMV^W122A‐GFP or pCamTVBMV^CPSTOP‐GFP, pCamTVBMVCP^WT or pCamTVBMVCP^W122A, and pBinP19. The OD₆₀₀ value of cultures carrying any plasmid was 0.6 before mixing. The expression of Nbef1a was used as an internal control. The TVBMV‐GFP (−)RNA accumulation was defined as 1. The results are presented as means ± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan's multiple range test (p < .05)

To determine whether TVBMVCP^WT and TVBMVCP^W122A could rescue the replication of TVBMV^W122A‐GFP, we infiltrated N. benthamiana leaves with a mixture of A grobacterium cultures harbouring pCamTVBMV^W122A‐GFP, pCamTVBMVCP^WT, or pCamTVBMVCP^W122A, and pBinP19, and analysed the harvested tissue samples through RT‐qPCR at 60 hpai. The accumulation level of TVBMV^W122A‐GFP (−)RNA in the leaf tissues coexpressing TVBMVCP^WT was partially rescued, whereas the accumulation level of TVBMV^W122A‐GFP (−)RNA in the leaf tissues coexpressing TVBMVCP^W122A was not changed (Figure 6c), indicating that TVBMVCP^WT could rescue the replication of TVBMV^W122A‐GFP.

We further found that TVBMVCP^WT and TVBMVCP^W122A could not rescue the replication and cell‐to‐cell movement of the replication‐deficient mutant TVBMV^CPSTOP‐GFP (Figure 6), suggesting that efficient viral replication is essential for TVBMV cell‐to‐cell movement.

2.7. The aromatic ring of W¹²² is crucial for CP stability, TVBMV replication, and movement

Several studies have shown that aromatic residues in proteins play essential roles in protein stability and function (Budyak et al., 2013; Butterfield et al., 2002; Chatterjee et al., 2019; Rege et al., 2018). To determine whether the aromatic ring of the residue W¹²² was responsible for TVBMV cell‐to‐cell movement and CP stability, we mutated the codon for residue W¹²² in pCamTVBMV‐GFP to the codons for the aromatic residue phenylalanine (F) or tyrosine (Y) to produce pCamTVBMV^W122F‐GFP (TVBMV^W122F‐GFP) and pCamTVBMV^W122Y‐GFP (TVBMV^W122Y‐GFP), and then individually infiltrated them into N. benthamiana leaves. By 7 dpai, both mutants TVBMV^W122F‐GFP and TVBMV^W122Y‐GFP caused systemic infection in N. benthamiana, although the GFP fluorescence from these two mutants was weaker than that of TVBMV‐GFP under a UV light (Figure 7a). Western blot results revealed that the two mutants’ CP accumulated approximately 82% and 77%, respectively, of the TVBMV‐GFP CP level (Figure 7b). Analysis using RT‐qPCR showed that, at 60 hpai, the accumulation levels of viral RNA in the systemic leaves and (−)RNA in the TVBMV^W122F‐GFP and TVBMV^W122Y‐GFP‐inoculated N. benthamiana leaf patches were approximately 60% of the TVBMV‐GFP level (Figure 7c,d).

FIGURE 7.

Effects of aromatic residues on TVBMV infection and CP stability. (a) GFP fluorescence in the systemic leaves of Nicotiana benthamiana plants infected with TVBMV‐GFP, TVBMV^W122F‐GFP, and TVBMV^W122Y‐GFP at 7 days postagroinfiltration (dpai). The W¹²² codon in pCamTVBMV‐GFP was mutated to the F or Y codons to produce pCamTVBMV^W122F‐GFP and pCamTVBMV^W122Y‐GFP, which were individually infiltrated into N. benthamiana leaves. (b) and (c) The accumulation levels of TVBMV CP and RNA in different leaf samples at 7 dpai. Accumulation of CP was determined through western blot assay using a TVBMV CP‐specific antibody. Ab, antibody. Coomassie brilliant blue‐stained RuBisCO large subunit protein (Rubi) was used to show sample loadings. The expression of Nbef1a was used as an internal control in quantitative reverse transcription PCR (RT‐qPCR). (d) The accumulation levels of TVBMV (−)RNA in N. benthamiana leaf patches infiltrated with pCamTVBMV‐GFP, pCamTVBMV^W122F‐GFP, or pCamTVBMV^W122Y‐GFP at 60 hpai were analysed using RT‐qPCR. The expression of Nbef1a was used as an internal control in RT‐qPCR. The results are presented as means ± SD from three biological replicates per treatment. (e) GFP fluorescence in N. benthamiana leaf patches agroinfiltrated with pCamGFP‐TVBMVCP, pCamGFP‐TVBMVCP^W122A, pCamGFP‐TVBMVCP^W122F, and GFP‐TVBMVCP^W122Y. The leaf was photographed under a UV light at 4 dpai. (f) Relative GFP fluorescence intensity in N. benthamiana leaf patches agroinfiltrated with pCamGFP‐TVBMVCP, pCamGFP‐TVBMVCP^W122A, pCamGFP‐TVBMVCP^W122F, and GFP‐TVBMVCP^W122Y. Analysis of GFP fluorescence intensity was conducted at 4 dpai using a microplate reader. The results are presented as means ± SD from three biological replicates per treatment. The statistical significance between treatments was determined using Duncan's multiple range test (p < .05)

We then mutated the W¹²² codon in pCamGFP‐TVBMVCP to produce pCamGFP‐TVBMVCP^W122F (GFP‐TVBMVCP^W122F) and pCamGFP‐TVBMVCP^W122Y (GFP‐TVBMVCP^W122Y), and individually infiltrated them into N. benthamiana leaves. At 4 dpai, the GFP fluorescence intensity from GFP‐TVBMVCP^W122F and GFP‐TVBMVCP^W122Y reached approximately 60% of GFP‐TVBMVCP, whereas that from GFP‐TVBMVCP^W122A was only 17% (Figure 7e,f). When the codon for residue A in pCamGFP‐TVBMVCP^W122A was changed back to the codon for residue W, the reverted mutant GFP‐TVBMVCP^W122A‐W produced similar GFP fluorescence as GFP‐TVBMVCP (Figure S6a,b). In this study, we also mutated the codon for W¹²² in pCamGFP‐TVBMVCP to codons for nonaromatic residues R, asparagine (N), D, C, E, glutamine (Q), glycine (G), H, isoleucine (I), leucine (L), K, methionine (M), proline (P), S, threonine (T), or valine (V). After the infiltration of these mutant constructs to N. benthamiana leaves, none of the mutant fusion proteins produced visible GFP fluorescence in the leaf patches (Figure S6a,b). When the above mutations were introduced into pCamTVBMV‐GFP, all the mutant viruses were confined to single cells (Figure S6c). Deletion of the W¹²² codon from pCamGFP‐TVBMVCP or pCamTVBMV‐GFP reduced the accumulation level of the mutant GFP‐TVBMVCP^W122A and abolished the cell‐to‐cell movement of TVBMV ^W122del‐GFP (Figure S6).

2.8. The corresponding residues of W¹²² in watermelon mosaic virus CP or potato virus Y CP are also critical for viral cell‐to‐cell movement and CP accumulation

W¹³³ in watermelon mosaic virus (WMV) CP and residue W¹¹⁸ in potato virus Y (PVY) CP are the corresponding residues of W¹²² in TVBMV CP. To investigate whether W¹³³ and W¹¹⁸ were also crucial for viral cell‐to‐cell movement, we deleted the codon for W¹³³ in pCBWMV‐GFP and the codon for W¹¹⁸ in pCamPVY‐GFP to produce pCBWMV^W133del‐GFP and pCamPVY^W118del‐GFP, respectively. We then individually infiltrated pCBWMV‐GFP, pCBWMV^W133del‐GFP, pCamPVY‐GFP, and pCamPVY^W118del‐GFP into N. benthamiana leaves. At 132 hpai, GFP fluorescence was observed in clusters of multiple cells infected with WMV‐GFP or PVY‐GFP. In contrast, GFP fluorescence from WMV^W133del‐GFP or PVY^W118del‐GFP was confined to single cells (Figure 8a,b), indicating that residues W¹³³ and W¹¹⁸ were also crucial for viral cell‐to‐cell movement.

FIGURE 8.

Deletion of W¹³³ from watermelon mosaic virus coat protein (WMV CP) or W¹¹⁸ from potato virus Y (PVY) CP abolished viral cell‐to‐cell movement and reduced viral CP accumulation. (a) The cell‐to‐cell movement of WMV‐GFP, WMV^W133del‐GFP, PVY‐GFP, and PVY^W118del‐GFP in Nicotiana benthamiana plants at 132 hr postagroinfiltration. The infiltrated leaves were sampled, examined, and imaged under a confocal microscope. GFP fluorescence indicates a virus infection. (b) The numbers of cells with GFP fluorescence presented as the means ± SD from 30 infection foci per treatment. The infection foci caused by WMV‐GFP or PVY‐GFP infection had too many cells with GFP fluorescence and are not presented here. (c) GFP fluorescence in the leaf patches expressing GFP‐WMVCP, GFP‐WMVCP^W133del, GFP‐PVYCP, or GFP‐PVYCP^W118del. (d) GFP fluorescence intensity in different leaf patches expressing GFP‐WMVCP, GFP‐WMVCP^W133del, GFP‐PVYCP, or GFP‐PVYCP^W118del as determined at 4 days postagroinfiltration using a microplate reader. The results are presented as means ± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan's multiple range test (p < .05)

To investigate the roles of W¹³³ and W¹¹⁸ in viral CP accumulation, we constructed pCamGFP‐WMVCP, pCamGFP‐WMVCP^W133del, pCamGFP‐PVYCP, and pCamGFP‐PVYCP^W118del, and then individually infiltrated them into N. benthamiana leaves. The results showed that, at 4 dpai, the GFP fluorescence from the two mutant proteins was significantly weaker than that from the two parental proteins (Figure 8c,d), indicating that residue W¹³³ in WMV CP and residue W¹¹⁸ in PVY CP are crucial for CP accumulation.

3. DISCUSSION

We studied the role of the conserved residue W¹²² in the TVBMV CP core domain in viral cell‐to‐cell movement and replication. Our results showed that mutation of W¹²² to various nonaromatic residues reduced viral replication and CP stability, leading to a defective viral cell‐to‐cell movement. Additionally, the aromatic ring of W¹²² was a key determinant of TVBMV CP stability, viral replication, and cell‐to‐cell movement.

CP is one of the potyviral proteins involved in viral cell‐to‐cell movement. It has been reported that a change in a single residue in the CP can affect potyviral cell‐to‐cell movement. For example, mutation of conserved residue S¹²⁹, R¹⁶¹, or D²⁰⁵ (positions are numbered according to the TVBMV CP sequence) in the TEV CP core domain abolishes viral cell‐to‐cell movement (Dolja et al., 1994, 1995). In this study, the aromatic residue W¹²² in the CPs of 139 potyviruses was completely conserved (Figure S1). Mutation of W¹²² to various nonaromatic residues abolished TVBMV cell‐to‐cell movement (Figures 1e,f and S6c). Additionally, the deletion of W¹³³ residue from WMV CP and the W¹¹⁸ residue from PVY CP abolished WMV and PVY cell‐to‐cell movement in N. benthamiana leaves (Figure 8a,b). Because W¹¹⁸ and W¹³³ corresponded to W¹²² in TVBMV CP, we concluded that the conserved residue W¹²² was a key determinant of the potyviral cell‐to‐cell movement.

It is noteworthy that the three TEV movement‐deficient mutants derived from S¹²⁹, R¹⁶¹, and D²⁰⁵ do not produce virus particles (Dolja et al., 1994, 1995). The deletion of 65 residues from the C‐terminus of wheat streak mosaic virus (Tritimovirus, family Potyviridae) CP inhibits viral cell‐to‐cell movement but has no apparent effect on virus particle formation (Tatineni et al., 2014). We found that mutations of W¹²² to A, E, or K did not change TVBMV particle morphology, but inhibited viral cell‐to‐cell movement (Figures 1 and 4). Wei et al. (2010) reported that in the presence of turnip mosaic virus (TuMV) P3N‐PIPO or during TuMV infection, TuMV CP and CI can accumulate adjacent to PD, which is a necessary step during potyviral cell‐to‐cell movement. In this study, we discovered that TVBMV CP and CI also accumulated adjacent to the PD in the presence of P3N‐PIPO and TVBMV infection. Furthermore, the ability of TVBMV CP to target PD was not affected by the change in W¹²² to A (Figure 5). Therefore, we concluded that the role of W¹²² was in viral cell‐to‐cell movement, but not in viral RNA encapsidation.

Previous studies have shown that potyviral replication and cell‐to‐cell movement are coupled processes (Chai et al., 2020; Cui & Wang, 2016). Here, we found that substitution of conserved W¹²² to nonaromatic residues significantly inhibited, but did not abolish, the replication of mutant viruses (Figure 2c,d). Interestingly, the CP accumulation levels of these W¹²² virus mutants were similar to that of the defective virus mutant TVBMV^NIbΔGDD‐GFP (Figure 2b). The abolished cell‐to‐cell movement of TVBMV^W122A‐GFP could be rescued by coexpressing TVBMVCP^WT (Figure 6). In the same experiment, the RNA accumulation level of TVBMV^W122A‐GFP was also increased. These results indicate that the reduced viral replication and extremely low CP accumulation might be both responsible for the defective cell‐to‐cell movement of W¹²² mutant viruses.

Potyviruses adopt a polyprotein expression strategy. The accumulation of individual viral proteins in the same tissues is assumed to be at the same level. The accumulating levels of (+)RNA and (−)RNA derived from W¹²² mutant viruses were much higher than those of the TVBMV^NIbΔGDD‐GFP and TVBMV^CPSTOP‐GFP mutant viruses (Figure 2c,d). The GFP accumulation levels of the three W¹²² mutant viruses were significantly higher than that of TVBMV^NIbΔGDD‐GFP, but not the accumulation levels of their CPs (Figure 2a,b). We considered that the mutations introduced to W¹²² destabilized the CPs. This conclusion was supported because substitution of A for W¹²² in the polyprotein NIa:HA‐NIb:GFP:CP did not change the accumulation levels of HA‐NIb and GFP, but reduced the accumulation level of mutant CP (Figure 3c).

Ubiquitin signalling and autophagy are two critical pathways that control protein degradation (Goldberg, 2012; Olzmann et al., 2004, 2007). It has been reported that inhibition of proteasome activity using MG132 enhanced potato virus A CP accumulation (Lõhmus et al., 2017), indicating that ubiquitin signalling can regulate virus CP degradation. The conserved helix in the C‐terminal region of the NSs protein of the watermelon silver mottle virus (WSMoV) is critical for protein stability. However, the accumulation of WSMoV with mutations in the conserved helix is not affected by MG132 treatment (Huang et al., 2020). In this study, treatment of leaves with MG‐132 or 3‐MA, a specific inhibitor of autophagy (Seglen & Gordon, 1982), had no significant effect on the accumulation of GFP‐TVBMVCP or mutant GFP‐TVBMVCP^W122A (Figure S7), suggesting that ubiquitin signalling and autophagy may not be the factors causing TVBMV CP instability.

The aromatic residues W, F, and Y are known to be important for protein stability and function. For example, the aromatic ring of Y³⁹⁸ in the WSMoV NSs protein has been shown to affect NSs stability and its RNA silencing suppression activity (Huang et al., 2015). Deletion of five residues, including an aromatic residue, from the readthrough protein of potato leafroll virus affects viral systemic infection and disease symptom induction (Xu et al., 2018). In this study, after mutation of W¹²² to nonaromatic residue A, E, or K, the mutant viruses produced destabilized CPs and exhibited reduced viral replication and defective viral cell‐to‐cell movement (Figures 1 and 2). Substitutions of other nonaromatic residues for W¹²² produced the same results (Figure S6). In contrast, changing residue W¹²² to aromatic residues F or Y yielded two mutant viruses capable of replicating to c.60% of TVBMV‐GFP in the infiltrated leaves and c.80% of TVBMV CP in the systemically infected leaves (Figure 7d). Because similar results were obtained for WMV and PVY CPs (Figure 8), we concluded that the aromatic residue W¹²² was a determinant of potyviral CP stability, replication, and movement. Aromatic rings can interact with each other to stabilize proteins. In a recent report, aromatic ring interaction between Y⁵¹ and F⁶⁴ in the small ubiquitin‐like modifier (SUMO) is vital to SUMO stability and its SUMOylation activity (Chatterjee et al., 2019). We speculate that the aromatic ring of W¹²² also interacts with other residues to stabilize CP, leading to successful cell‐to‐cell movement.

The aromatic ring of residue W¹²² is a critical factor for potyviral replication and CP stability. As a result, mutation of W¹²² to a nonaromatic residue abolishes potyviral cell‐to‐cell movement. The findings presented here increase our understanding of the underlying mechanism controlling potyviral cell‐to‐cell movement. Further studies are necessary to elucidate the roles of CP secondary or higher structures during potyvirus infection in plants.

4. EXPERIMENTAL PROCEDURES

4.1. Amino acid sequence alignments

CP sequences of 139 potyviruses were downloaded from a reference sequence database at the National Center for Biotechnology Information (O’Leary et al., 2016). Multiple sequence alignments were performed using the ClustalW program in BioEdit v. 7.2.6, using the default parameters (Hall, 1999). The resulting FASTA file was further processed using the online application WebLogo (http://weblogo.berkeley.edu) to generate sequence logos as previously described (Crooks, 2004).

4.2. Plasmid construction and site‐directed mutagenesis

The gfp gene‐containing infectious clones pCamTVBMV‐GFP (accession number: JQ407082), pCBWMV‐GFP, pCamPVY‐GFP, and the infectious clone pCamTVBMV without gfp gene were constructed in our laboratory (Gao et al., 2012; Geng et al., 2015). The coding sequences of TVBMV CP, WMV CP, and PVY CP were PCR‐amplified from the above infectious clones and individually inserted into the expression vector pCam35S::GFP to produce pCamGFP‐TVBMVCP, pCamGFP‐WMVCP, and pCamGFP‐PVYCP, respectively. The TVBMV CI coding region was PCR‐amplified and cloned into a different expression vector pCam35S::DsRed to produce pCamCI‐DsRed. The coding regions of TVBMV P3N‐PIPO, CP, and NIaPro:NIb:GFP:CP were PCR‐amplified and cloned individually into the pCam35S vector to produce pCamTVBMVCP^WT, pCamP3N‐PIPO, and pCamNIaPro:NIb:GFP:CP, respectively. An HA‐tag encoding sequence was inserted between the first and second codons of the NIb open reading frame (ORF) in the pCamNIaPro:NIb:GFP:CP clone to produce pCamNIaPro:HA‐NIb:GFP:CP. For expression of TVBMV CP in E. coli cells, the CP sequence was cloned into vector pEHISTEV to produce pEHISTEV‐TVBMVCP. To generate a replication‐deficient TVBMV mutant, we deleted the codons for the conserved GDD motif in the NIb ORF in pCamTVBMV‐GFP to produce pCamTVBMV^NIbΔGDD‐GFP as described (Geng et al., 2017). A stop codon was inserted immediately after the second codon in the CP ORF in pCamTVBMV‐GFP to produce pCamTVBMV^CPSTOP‐GFP. Substitutions of the codons for alanine (A), glutamic acid (E), or lysine (K) for that of W¹²² in pCamTVBMV‐GFP were performed using site‐directed mutagenesis as previously described (Liu & Naismith, 2008) to produce pCamTVBMV^W122A‐GFP, pCamTVBMV^W122E‐GFP, and pCamTVBMV^W122K‐GFP, respectively. The codon for W¹³³ in pCBWMV‐GFP and the codon for W¹¹⁸ in pCamPVY‐GFP were deleted to produce pCBWMV^W133del‐GFP and pCamPVY^W118del‐GFP, respectively. All plasmids were sequenced before use. Primers used in this study are listed in Table S1. The names and products of various constructs used in this study are listed in Table S2.

4.3. Plant growth, virus inoculation, and protein transient expression

N. benthamiana plants were grown in a greenhouse set at 25 °C and a 16/8 hr (light/dark) photoperiod. Plasmids were individually transformed into Agrobacterium tumefaciens GV3101 cells using a freeze‐thaw method. Agrobacterium cultures harbouring different plasmids were cultured overnight in Luria–Bertani liquid medium supplemented with appropriate antibiotics at 28 °C with 220 rpm shaking. The cultures were pelleted through centrifugation at 8,000 × g for 2 min and individually resuspended in an induction buffer containing 10 mM MgCl₂, 10 mM MES, and 150 μM acetosyringone. To determine viral accumulation in the infiltrated leaf patches and systemic movement in the plants, the agrobacterium cultures were diluted to OD₆₀₀ = 0.2. To determine viral cell‐to‐cell movement, the agrobacterium cultures were diluted to OD₆₀₀ = 0.0001. For transient expression assays, individual agrobacterium cultures were diluted to OD₆₀₀ = 0.4 and then mixed with an agrobacterium culture (1:1 vol/vol) harbouring pBinP19, a vector expressing tomato bushy stunt virus P19 protein. For cell‐to‐cell movement trans‐complementation assays, the agrobacterium cultures carrying pCamTVBMV^CPSTOP‐GFP or pCamTVBMV^W122A‐GFP were diluted to OD₆₀₀ = 0.0003 and the cultures carrying pCamTVBMVCP^WT or pCamTVBMVCP^W122A were diluted to OD₆₀₀ = 0.6. These cultures were individually mixed with an equal volume of culture carrying pBinP19 (OD₆₀₀ = 0.6). For viral replication trans‐complementation assays, the agrobacterium cultures carrying pCamTVBMV^CPSTOP‐GFP or pCamTVBMV^W122A‐GFP, pCamTVBMVCP^WT or pCamTVBMVCP^W122A, and pBinP19 were diluted to OD₆₀₀ = 0.6 and mixed at a ratio of 1:1:1. After 3 hr incubation at 25 °C, the cultures were individually infiltrated into leaves of 4–6‐week‐old N. benthamiana plants with 1‐ml needleless syringes. The infiltrated plants were grown in a greenhouse till use. For protein degradation assays, a 1% dimethyl sulphoxide (DMSO) (control) solution or 50 μM MG132 (Selleck Chemicals) in 1% DMSO was infiltrated into N. benthamiana leaves to inhibit the ubiquitin signalling pathway. Water (control) or a 5 mM 3‐methyladenine (3‐MA) (Selleck Chemicals) aqueous solution was used to inhibit the autophagy pathway in leaves.

4.4. RNA extraction and RT‐qPCR

The infiltrated leaf tissues or the systemic leaves were collected from the assayed N. benthamiana plants and used for total RNA extraction using TransZol reagent (TransGen Biotech). The resulting total RNA samples were treated with a gDNA wipe enzyme (Vazyme) to remove plant genomic DNA. For RT‐qPCR analysis of viral RNA accumulation, 500 ng total RNA (per sample) was reverse transcribed using a HiScript II Q RT SuperMix kit supplemented with random primers (Vazyme). To detect viral plus‐ and minus‐strand RNA accumulation in the infiltrated leaf tissues, 500 ng total RNA (per sample) was reverse transcribed using primers qTVBMVCP‐R and EF1A‐R or primers qTVBMVCP‐F and EF1A‐R, and the HiScript II Q RT select SuperMix. Quantitative PCR was conducted using a ChamQ SYBR qPCR Master Mix (Vazyme) on a thermocycler (LC96; Roche). TVBMV CP‐specific qPCR primers are listed in Table S1. The expression of N. benthamiana ef1a was determined using primer EF1A‐F and EF1A‐R, and was used as an internal control.

4.5. Western blot assay

Systemically infected leaves or the infiltrated leaf tissues were collected from the assayed N. benthamiana plants and homogenized individually in a protein extraction buffer (100 mM Tris‐HCl, 150 mM NaCl, 1 mM EDTA, 5% sucrose, and 1 mM phenylmethanesulfonyl fluoride) at a ratio of 1:2 (wt/vol) using a tissue grinder (Jingxin). The leaf extracts were denatured at 95 °C for 5 min, incubated on ice for 5 min, and then centrifuged at 12,000 × g for 5 min. The supernatant was collected from each sample, and the proteins in each sample were separated in SDS‐polyacrylamide gels through electrophoresis and then blotted onto nitrocellulose membranes. Polyclonal antibodies specific for TVBMV CP or GFP were prepared in our laboratory (Ji et al., 2018; Lan et al., 2007) and were all used at a 1:1,000 (vol/vol) dilution. The HA‐specific antibody (Thermo Fisher Scientific) was used at a 1:1,000 dilution. A horseradish peroxidase‐conjugated goat anti‐rabbit IgG (Sigma‐Aldrich) was used as the secondary antibody diluted at 1:50,000 (vol/vol). After the addition of the SuperSignal West Dura extended duration substrate solution (Thermo Fisher Scientific), the detection signal was visualized using a chemiluminescent imaging and analysis system (Sage).

4.6. ELISA

The collected plant tissues were individually homogenized (1:8, wt/wt) in a coating buffer (15 mM Na₂CO₃, 35 mM NaHCO₃, pH 9.6). Aliquots (100 μl) of leaf crude extracts were added into individual wells on a 96‐well microtitre plate and incubated overnight at 4 °C. After four rinses with a phosphate‐buffered saline with Tween‐20 (PBS‐T; 80 mM Na₂HPO₄, 1.5 M NaCl, 20 mM KH₂PO₄, 30 mM KCl, 0.5% Tween‐20, pH 7.4), the GFP or the TVBMVCP antibody solution was added into each well, and the plate was incubated at 37 °C for 4 hr. After four rinses with the PBS‐T an alkaline phosphatase‐conjugated goat anti‐rabbit IgG solution was added to the wells followed by 4 hr incubation at 37 °C. After the addition of a p‐nitrophenyl phosphate substrate solution (Sigma, 0.25 mg/ml), the absorbance value (A ₄₀₅) of each well was measured using a microplate reader (BioTek Synergy Mx).

4.7. Confocal microscopy and fluorescence intensity measurement

To monitor viral intercellular movement and protein subcellular localization in N. benthamiana leaves, the agrobacterium‐infiltrated leaf patches were collected and examined under a laser confocal microscope (Carl Zeiss). For GFP fluorescence observation, the excitation and emission wavelengths were set at 488 and 520–540 nm, respectively. For DsRed fluorescence observation, the excitation and emission wavelengths were set at 561 and 590–630 nm, respectively. The captured images were processed using the ZEN 2.1 (Carl Zeiss). Agrobacterium‐infiltrated N. benthamiana leaves were photographed under UV light (365 nm) from a hand‐held UV lamp (LUYOR) using a digital camera (Canon 80D). To determine GFP fluorescence intensity in the infiltrated leaf tissues, leaf discs (5 mm in diameter) were sampled from the infiltrated N. benthamiana leaf patches with a cork borer and individually placed in wells of a 96‐well microtitre plate. GFP fluorescence from each well was determined using a microplate reader (BioTek). The excitation wavelength was 485/10 nm and the emission wavelength was 535/10 nm.

4.8. Virus particle purification

Virus particles were purified from the infiltrated N. benthamiana leaves. At 5 dpai, 15 g of tissues were harvested from the infiltrated leaves, ground in liquid nitrogen, and then homogenized in 30 ml of 0.2 M phosphate buffer at pH 8 and supplemented with 0.15% β‐mercaptoethanol and 0.01 M EDTA. The crude extracts were centrifuged at 8,000 × g for 20 min. The supernatant was filtered through four layers of cheesecloth and stirred at 4 °C for 3 hr after adding 1% Triton X‐100, 40 g/L polyethylene glycol 6,000, and 0.2 M NaCl. Virus particles were precipitated by centrifugation at 8,000 × g for 20 min, and the pellets were resuspended overnight at 4 °C in a 0.2 M phosphate buffer at pH 8 and containing 1% Triton X‐100. Insoluble materials were removed by centrifugation at 8,000 × g for 20 min, and virus particles in the supernatant were pelleted through 1 hr ultracentrifugation at 100,000 × g at 4 °C using an ultracentrifuge CP100WX (Hitachi). The pellets were individually resuspended overnight in a 0.05 M phosphate buffer at pH 8 and at 4 °C. The insoluble materials were removed again by 20 min centrifugation at 8,000 × g. Small amounts of supernatant from each sample were negatively stained with 2% uranyl acetate, loaded onto 230‐mesh carbon‐coated copper grids, and then examined under a JEM‐1200Ex transmission electron microscope (Jeol) for virus particle morphology.

4.9. Protein expression

Plasmids for expressing the wild‐type or mutant CPs were individually transformed into E. coli Rosetta cells, and the transformed cells were cultured until OD₆₀₀ = 0.6. The cultures were induced with a 0.1 mM isopropyl‐β‐d‐thiogalactopyranoside solution and incubated at 16 °C for 12 hr. The cultures were pelleted by centrifugation at 8,000 × g for 5 min, and the pellets were individually resuspended with a lysis buffer (50 mM Na₂HPO₄, 300 M NaCl, pH 8.0) followed by 5 min of lysis through sonication on ice. The lysed cells were pelleted by centrifugation at 10,000 × g for 5 min. Supernatants of each sample were transferred into a new clean tube and used as soluble proteins for SDS‐PAGE and western blot assays.

Supporting information

FIGURE S1 Alignment of CP sequences of 139 potyviruses. The start and end positions of the CP core domain are indicated with red arrowheads. The positions of residues R⁶², S⁹², and K²⁴⁶ in the TVBMV CP are indicated with blue arrows, whereas the position of residue W¹²² is indicated with a red arrow. Blue, red, and black letters are used to indicate the positively charged, negatively charged, and neutral residues, respectively

FIGURE S2 Mutations of nonconserved residues R⁶², S⁹², and K²⁴⁶ in the TVBMV CP core domain did not inhibit virus systemic infection. Nicotiana benthamiana plants were inoculated with TVBMV‐GFP, TVBMV^R62A‐GFP, TVBMV^R62C‐GFP, TVBMV^R62E‐GFP, TVBMV^S92A‐GFP, TVBMV^S92D‐GFP, TVBMV^S92H‐GFP, TVBMV^K246A‐GFP, TVBMV^K246C‐GFP, or TVBMV^K246E‐GFP through agroinfiltration. Viral systemic infection in these assayed plants was determined by the accumulation of CP through ELISA at 7 days post agroinfiltration. The results are presented as means ± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan’s multiple range test (p < .05)

FIGURE S3 Accumulation of TVBMV CP and GFP in the infiltrated Nicotiana benthamiana leaves at 108 hr post agroinfiltration (hpai). N. benthamiana leaves were inoculated with TVBMV‐GFP, TVBMV^W122A‐GFP, TVBMV^W122E‐GFP, TVBMV^W122K‐GFP, TVBMV^NIbΔGDD‐GFP, or TVBMV^CPSTOP‐GFP through agroinfiltration. The accumulation of TVBMV CP (a) and GFP (b) in the infiltrated leaves was determined at 108 hpai through ELISA. The results are presented as means± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan’s multiple range test (p < .05)

FIGURE S4 Detection of TVBMVCP^WT and TVBMVCP^W122A in the infiltrated Nicotiana benthamiana leaves. TVBMVCP^WT and TVBMVCP^W122A were expressed individually in N. benthamiana leaves through agroinfiltration. The accumulating levels of TVBMVCP^WT and TVBMVCP^W122A in the leaf patches were determined through western blot assay. Ab, antibody. The Coomassie brilliant blue‐stained RuBisCO large subunit protein (Rubi) was used to show sample loadings

FIGURE S5 Expression of wild‐type and mutant CP of TVBMV in Escherichia coli cells. E. coli Rosetta cells were transfected with pEHISTEV‐TVBMVCP or pEHISTEV‐TVBMVCP^W122A. The transfected E. coli cells were cultured, induced at 16 °C for 12 hr, and analysed for protein expressions through western blot assay. The upper panel shows the accumulation of CP (arrowhead) stained with Coomassie brilliant blue and the lower panel shows CP (arrowhead) detected by western blot. T, total protein; S, soluble proteins; NS, nonsoluble proteins

FIGURE S6 Effects on substitution of nonaromatic residues for W¹²² on TVBMV CP stability and viral cell‐to‐cell movement. (a) The W¹²² codon in pCamGFP‐TVBMVCP was mutated to the codons for each of the 17 nonaromatic residues. Additionally, the codon for residue A in pCamGFP‐TVBMVCP^W122A was changed back to the codon for W to produce pCamGFP‐TVBMVCP^W122A‐W. These plasmids were individually agroinfiltrated into Nicotiana benthamiana leaves. (a) The infiltrated leaves were photographed at 4 days postagroinfiltration (dpai). (b) Green fluorescent protein (GFP) fluorescence intensity in various leaf tissues was determined using a microplate reader at 4 dpai. The GFP fluorescence intensity in the tissue expressing GFP‐TVBMVCP is considered to be 1. A W¹²² deletion mutant, pCamGFP‐TVBMVCP^W122del, was also included in this experiment. The results are presented as means ± SD from three biological replicates per treatment. (c) The W¹²² codon in pCamTVBMV‐GFP was mutated to the codons for each of the 15 nonaromatic residues. These plasmids were individually agroinfiltrated into N. benthamiana leaves. Cell‐to‐cell movement of these mutants was observed under a confocal microscope. The numbers of cells with GFP fluorescence per foci per treatment were recorded at 132 hr post agroinfiltration (hpai). The values are presented as means ± SD from 30 infection foci per treatment. Statistical significance between treatments was determined using Duncan’s multiple range test (p < .05)

FIGURE S7 CP instability was not associated with an autophagy‐ or ubiquitin‐related pathway. (a) and (b) Nicotiana benthamiana leaves were agroinfiltrated with pCamGFP‐TVBMVCP or pCamGFP‐TVBMVCP^W122A, and then with MG‐132 or DMSO. The infiltrated leaves were photographed at 12 and 36 hr post chemical treatment (hpct) to determine the effect on autophagy. (c) and (d) Leaves infiltrated with 3‐MA or water were also photographed to show the impact of the ubiquitin pathway. (e)–(h) The intensity of green fluorescent protein (GFP) fluorescence in various tissue samples was determined using a microplate reader. The results are presented as means ± SD from three biological replicates per treatment. Statistical significance between treatments was determined using Duncan’s multiple range test (p < .05)

TABLE S1 Primers used in this study

TABLE S2 Names and products of various constructs used in this study

Yan Z‐Y, Cheng D‐J, Liu L‐Z, et al. The conserved aromatic residue W¹²² is a determinant of potyviral coat protein stability, replication, and cell‐to‐cell movement in plants. Mol Plant Pathol. 2021;22:189–203. 10.1111/mpp.13017

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.