Both Conserved Aromatic Nature and Critical RNA Motif of Residue 67 in the ToBRFV Movement Protein May Be Essential for Evading Tm‐2² ‐Mediated Resistance

ABSTRACT

The 68th amino acid residue (cysteine, C68) in tobacco mosaic virus (TMV) movement protein (MP) and the corresponding position in some tobamoviruses are required to trigger Tm‐2 ² ‐mediated resistance. In contrast, the 67th amino acid residue (histidine, H67) in tomato brown rugose fruit virus (ToBRFV) MP enables evasion, although its key property remains unclear. To assess the key property of this site, a saturated mutant pool was generated at ToBRFV MP H67 and TMV MP C68, along with two single mutants, H67C and C68H. Infection assays were conducted using two tomato cultivars, Jingpeng No. 1 (JP, Tm‐2 ² /tm‐2) and Rutgers (RG, tm‐2/tm‐2), as well as wild‐type (WT) and Tm‐2 ² transgenic Nicotiana benthamiana. Progeny sequencing revealed species‐ and cultivar‐specific outcomes. The mutants TAC and TAT (encoding the aromatic amino acid tyrosine, Y) and CAC (encoding the WT histidine, H), were the only ones detected in JP tomato and Tm‐2 ²  N. benthamiana. In single‐infection assays, ToBRFV carrying TTC (encoding phenylalanine, F) at this position infected JP tomato, TAC (Y) enabled infection of Tm‐2 ²  N. benthamiana. TGG (encoding tryptophan, W) infected both JP tomato and Tm‐2 ²  N. benthamiana. Moreover, CAT (H) but not CAC (H), and TTC (F) but not TTT (F) infected JP plants, highlighting the role of the critical RNA motif. This suggests that the aromatic moiety and critical RNA motif at this position are essential for evading Tm‐2 ² ‐mediated resistance, as the imidazole ring of the WT histidine also exhibits aromatic character. Therefore, Tm‐2 ² may be engineered to recognize aromatic amino acids at this position to confer resistance to ToBRFV.

Saturated mutagenesis of the MP C/H site reveals that residue identity at this position determines tobamovirus infection and Tm‐2 ² resistance evasion in tomato cultivars.

1. Introduction

Plant viruses represent a persistent and significant threat to global food security, causing substantial losses in crop yield and quality (Tatineni and Hein 2023). Tomato plants possess several resistance genes against tobamoviruses, with Tm‐2 and Tm‐2 ² being key examples (Qian et al. 2018; Sánchez‐Sánchez et al. 2023; Tanwir 2012). The Tm‐2 ² gene, in particular, has served as a cornerstone of resistance against tobamoviruses for over 60 years. This resistance is mediated by a nucleotide‐binding leucine‐rich repeat (NLR) immune receptor encoded by Tm‐2 ² , which recognizes the viral movement protein (MP) as an effector (Hak et al. 2023; Rivera‐Márquez et al. 2022). Previous studies have identified amino acid substitutions in the movement protein (MP) of tomato mosaic virus (ToMV) that enable the virus to overcome Tm‐2 ² ‐mediated resistance. For example, three substitutions, Arg‐238, Tyr‐240 and Glu‐244, located in the carboxy‐terminal region of the ToMV MP, were shown to confer the ability to replicate in Tm‐2 ² tomato plants (Kuroiwa et al. 2022; Weber and Pfitzner 1998; Weber et al. 1993), highlighting the potential emergence of novel infectious viral strains capable of evading this resistance. The emergence of tomato brown rugose fruit virus (ToBRFV), which can overcome Tm‐2 ² ‐mediated resistance, has provided clear evidence supporting this notion (Hak and Spiegelman 2021). Unlike ToMV resistance‐breaking mutants, ToBRFV‐infected plants do not exhibit typical necrotic symptoms (Luria et al. 2017), suggesting that ToBRFV effectively evades Tm‐2 ² recognition rather than simply overcoming it through high‐level replication. Current hypotheses suggest that this evasion may involve not only mutations in the MP but also alterations in the charge of the replication enzyme and changes in disulphide bonds within the helicase domain (Maayan et al. 2018). Additionally, evolutionary adaptations and specific codon usage bias patterns in ToBRFV may further contribute to its ability to bypass established resistance genes (Ghorbani 2024). Thus, ToBRFV resistance‐breaking may involve multiple mechanisms.

It has been shown that a conserved cysteine (C68) in the movement protein (MP) of tobacco mosaic virus (TMV) is essential for both Tm‐2 ² activation and viral movement (Hak et al. 2023). Substitution of the ToBRFV MP residue H67 with the corresponding TMV MP C68 is sufficient to activate Tm‐2 ² ‐mediated resistance (Hak et al. 2023). In addition, several residues within the ToBRFV MP, including H67, N125, K129, A134, I147 and I168, have been identified as critical for overcoming Tm‐2 ² ‐mediated resistance (Yan et al. 2021). Recent studies showed that single amino acid substitutions in the ToBRFV MP can readily overcome resistance in newly developed tomato cultivars (Wang et al. 2025; Zisi et al. 2024). Despite these advances, the precise molecular mechanisms by which individual MP residues enable ToBRFV to evade Tm‐2 ² resistance remain poorly understood. This study therefore focused on the H67 position of the ToBRFV MP. By generating a saturated mutational library at this position and performing infection assays in tomato and Nicotiana benthamiana plants with different genetic backgrounds, particularly with respect to the presence of Tm‐2 ² , the key property of residue 67 required for resistance evasion was investigated. The results indicate that the conserved aromatic nature of residue 67 in the ToBRFV MP and the critical RNA motif at the corresponding region of the genome are important for evading Tm‐2 ² ‐mediated resistance. These findings provide a foundation for future studies aimed at developing novel and durable anti‐ToBRFV strategies.

2. Results

2.1. The Infection of ToBRFV and TMV Mutant Pools and Single Mutants in Two Tomato Cultivars

The reported 32 tobamoviruses can be divided into at least eight subgroups based on their host plants and a phylogenetic tree reported in a previous study, including Solanaceae‐, Brassicaceae‐, Cactaceae‐, Apocynaceae‐, Cucurbitaceae‐, Malvaceae‐, Leguminosae‐ and Passifloraceae‐infecting groups (Li et al. 2017). The amino acid sequences of the MP from 32 tobamoviruses were aligned. The corresponding accession numbers are listed in Table S1. In the Solanaceae‐infecting subgroup, the position corresponding to H67 in ToBRFV MP is occupied by cysteine (C) in 10 out of 11 viruses. The only exception is that asparagine (N) is present at the corresponding position in yellow tailflower mild mottle virus (YTMMV). In the Brassicaceae‐infecting subgroup, glycine (G) was found at the corresponding position in all six viruses. In the Cactaceae‐infecting subgroup, tyrosine (Y) was found at the corresponding position in the only two members. Threonine (T) was found in the only member of the Apocynaceae‐infecting subgroup, as well as in all five members of the Cucurbitaceae‐infecting subgroup. In the Malvaceae‐infecting subgroup, histidine (H) was found in both members. The Leguminosae‐infecting subgroup also contains two members, in which the corresponding position is occupied by aspartate (D). In the two viruses of Passifloraceae‐infecting subgroup, methionine (M) was found at the corresponding position. These residues are highlighted in a red box (Figure 1a). Mutant pools containing all possible nucleotide combinations of the ToBRFV MP H67 and TMV MP C68 regions were generated by PCR using degenerate primers. The library preparation method was described in a previous study (Wu et al. 2024). The pools were then sequenced by Sanger sequencing, revealing four peaks of similar height at each of the three positions, corresponding to the four nucleotides (Figure 1b). Over 50 clones from each pool were sequenced, and all amino acid mutants were detected. To further confirm that all codons were included in the mutant pool, deep sequencing was performed. The results showed that all 63 mutants and the wild type (WT) were present in both the ToBRFV MP H67 pool and the TMV MP C68 pool (Table S2). Two single mutants, ToBRFV MP H67C and TMV MP C68H, were generated through PCR. These single mutants, along with the two mutant pools, were subjected to infection assays using both Jingpeng No. 1 (JP) and Rutgers (RG) tomato cultivars, with four plants included in each experiment.

FIGURE 1.

Infection of tomato brown rugose fruit virus (ToBRFV) movement protein (MP) H67 mutant pool, tobacco mosaic virus (TMV) MP C68 mutant pool, and the ToBRFV MP H67C and TMV MP C68H mutants in Jingpeng No. 1 (JP) and Rutgers (RG) tomato cultivars. (a) Comparison of amino acid sequences of MPs from 32 tobamoviruses. The conserved cysteine (C) residue in some tobamoviruses and the histidine (H) in ToBRFV are highlighted in red‐dashed rectangles. (b) Preparation of ToBRFV MP H67 mutant pool and TMV MP C68 mutant pool. Degenerate primers were used to generate saturated mutant pools for the genomic regions corresponding to ToBRFV MP H67 and TMV MP C68. The pools were confirmed by Sanger sequencing, showing four peaks of similar height corresponding to the four nucleotides at each site of the three positions. Western blot analysis was used to assess infection of the mutant pools and single mutants (ToBRFV MP H67C and TMV MP C68H) in JP and RG tomato cultivars, with detection of coat protein (CP) at 2 (c) and 4 (d) weeks post‐inoculation (wpi). CBB, Commassie Briliant Blue loading control. (e) Symptoms of JP and RG tomato plants infected with mutant pools or single mutants at 6 wpi, Scale bar = 7 cm. M, mock; P, positive control; WT, wild type.

Western blot analysis was performed to detect ToBRFV coat protein (CP) at 2 weeks (Figure 1c) and 4 weeks post‐inoculation (wpi) (Figure 1d). At 2 wpi, the ToBRFV MP H67 pool infected one out of four Jingpeng No. 1 (JP) tomato plants, while the ToBRFV MP H67C mutant failed to infect any plants. In contrast, the ToBRFV MP H67 pool infected three out of four Rutgers (RG) plants, and ToBRFV MP H67C infected two out of four RG plants. The TMV C68H mutant pool failed to infect any JP plants but infected all four RG plants. Similarly, the TMV C68H single mutant did not infect any JP plants but infected three out of four RG plants (Figure 1c). At 4 wpi, the ToBRFV MP H67 pool still infected only one out of four JP plants, and the ToBRFV MP H67C mutant failed to infect any plants. The ToBRFV MP H67 pool continued to infect three out of four RG plants, while the ToBRFV MP H67C mutant infected all four RG plants. The TMV C68H mutant pool failed to infect any JP plants but infected all four RG plants. Similarly, the TMV C68H single mutant did not infect any JP plants but infected all four RG plants (Figure 1d). Symptoms in JP and RG tomato plants infected with mutant pools or single mutants at 6 wpi are presented (Figure 1e). More detailed symptoms were presented in Figure S1.

Progeny sequences of ToBRFV (Table 1) and TMV (Table 2) isolated from the JP and RG tomato cultivars, which were infected with the ToBRFV MP H67 mutant pool or the TMV MP C68 mutant pool, were sequenced at 2, 4 and 6 wpi. The ToBRFV MP H67 mutant pool infected only one JP plant, from which 10 clones were sequenced at 2 and 4 wpi; all clones showed the codon for tyrosine (Y) at the relevant position. No infection was detected at 6 wpi. Because only one plant became infected, the infection experiment was repeated; however, again only a single Tm‐2 ² plant was infected by ToBRFV H67 pool, and infection was detected only at 2 wpi (Figure S2). No infection was detected at 4 or 6 wpi. At 2 wpi, all 10 progeny clones carried the tyrosine (Y) codon (Table 1). In RG plants, the pool infected three plants. In RG plant 1, 10, 9 and 24 clones were sequenced at 2, 4 and 6 wpi, respectively. At 2 wpi, nine clones encoded cysteine (C) and one histidine (H); at 4 wpi, two encoded cysteine (C) and seven histidine (H); and at 6 wpi, all 24 clones encoded histidine (H). Notably, the codon for histidine here was not the WT CAT, but rather CAC. In RG plant 2, 10 and 7 clones were sequenced at 2 and 4 wpi, respectively, while no infection was detected at 6 wpi. All clones encoded phenylalanine (F). Similarly, in RG plant 3, 10 and 11 clones were sequenced at 2 and 4 wpi, respectively, with no infection detected at 6 wpi. All clones in this plant also encoded phenylalanine (F). However, the codon for histidine in plant 2 (TTC) was different from that in plant 3 (TTT) (Table 1). Progeny of ToBRFV H67C in RG plants were also sequenced. In all cases, the introduced mutation was maintained.

TABLE 1.

Progeny sequences of tomato brown rugose fruit virus (ToBRFV) isolated from Jingpeng No. 1 (JP) and Rutgers (RG), two tomato cultivars infected with ToBRFV movement protein (MP) H67 mutant pool.

Time (weeks post‐inoculation)	JP		RG
	Plant 1	Plant 2 (repeated experiment)	Plant 1	Plant 2	Plant 3
2	10 TAC (Y)	10 TAC (Y)	9 TGT (C), 1 CAC (H)	10 TTC (F)	10 TTT (F)
4	10 TAC (Y)	NA	2 TGT (C), 7 CAC (H)	7 TTC (F)	11 TTT (F)
6	NA	NA	24 CAC (H)	NA	NA

Note: The number of clones and the genome sequences at the H67 position are presented, with the corresponding amino acid sequence highlighted in red.

Abbreviation: NA, not available.

TABLE 2.

Progeny sequences of tobacco mosaic virus (TMV) isolated from Rutgers (RG) tomato infected with the TMV movement protein (MP) C68 mutant pool.

Time (weeks post‐inoculation)	RG
	Plant 1	Plant 2	Plant 3
2	7 TGT (C),1 AAT (N), 1 TTG (L)	7 TGC (C)	5 ACT (T), 5 TGG (W), 1 TCG (S)
4	10 TGT (C)	9 TGC (C), 8 TGT (C), 1 TCT (S)	2 ACT (T), 2 AGT (S), 2 AAC (N), 1 TGT (C), 1 TGC (C)
6	NA	8 TGT (C)	7 TGT (C)

Note: The number of clones and the genome sequences at the C68 position are presented, with the corresponding amino acid sequence highlighted in red.

Abbreviation: NA, not available.

The TMV MP C68 mutant pool did not infect any JP plants but infected three RG plants. In RG plant 1, 9 and 10 clones were sequenced at 2 and 4 wpi, respectively, with no infection detected at 6 wpi. At 2 wpi, seven clones encoded cysteine (C), one encoded asparagine (N) and one encoded leucine (L). At 4 wpi, all 10 clones encoded cysteine (C). In RG plant 2, 7, 18 and 8 clones were sequenced at 2, 4 and 6 wpi, respectively. At 2 wpi, all seven clones encoded cysteine (C); at 4 wpi, 17 clones encoded cysteine (C) and one encoded serine (S); at 6 wpi, all eight clones encoded cysteine (C). In RG plant 3, 11, 8 and 7 clones were sequenced at 2, 4 and 6 wpi, respectively. At 2 wpi, five clones encoded threonine (T), five encoded tryptophan (W) and one encoded serine (S); at 4 wpi, two clones encoded threonine (T), two encoded serine (S), two encoded asparagine (N) and two encoded cysteine (C); at 6 wpi, all seven clones encoded cysteine (C). Notably, different nucleotide codes for the same amino acids were frequently observed. The diversity in both nucleotide sequences and amino acids was higher in RG plants infected with the TMV pool than in plants infected with the ToBRFV pool (Table 2). Progeny of TMV C68H in RG plants were also sequenced. In all cases, the introduced mutation was maintained.

2.2. The Infection of ToBRFV and TMV Mutant Pools and Single Mutants in WT and Tm‐2 ² ‐Transgenic N. benthamiana

The genomes of JP (Tm‐2 ² /tm‐2) and RG (tm‐2/tm‐2) tomato cultivars differ at numerous loci (Zhang et al. 2026). To investigate the role of the Tm‐2 ² resistance gene in the infection by mutant pools and single mutants, we repeated the infection assay using WT N. benthamiana and N. benthamiana plants carrying a Tm‐2 ² transgene. Four plants were included in each experiment. Infection was confirmed by western blot at 10 days post‐infection (dpi). The ToBRFV pool infected all four Tm‐2 ² and WT plants, whereas the ToBRFV MP H67C mutant did not infect any Tm‐2 ² plants but infected all four WT plants. Similarly, the TMV pool and TMV MP C68H mutant did not infect any Tm‐2 ² plants but infected all WT plants (Figure 2a). Symptoms were recorded at 20 dpi (Figure 2b).

FIGURE 2.

Infection of tomato brown rugose fruit virus (ToBRFV) movement protein (MP) H67 mutant pool, tobacco mosaic virus (TMV) MP C68 mutant pool, and the ToBRFV MP H67C and TMV MP C68H mutants in wild‐type (WT) Nicotiana benthamiana and N. benthamiana with Tm‐2 ² transgene. (a) The infection of two mutant pools and two single mutants (ToBRFV MP H67C and TMV MP C68H) in Tm‐2 ² transgenic N. benthamiana and WT N. benthamiana was detected by western blot analysis of coat protein (CP) at 10 days post‐inoculation (dpi). RBL, RuBisCO loading control. (b) Symptoms in WT N. benthamiana and N. benthamiana with the Tm‐2 ² transgene, infected with mutant pools or single mutants, were observed at 20 dpi. M, mock; P, positive control.

Progeny of ToBRFV (Table 3) and TMV (Table 4) isolated from Tm‐2 ² and WT N. benthamiana plants infected with the ToBRFV MP H67 mutant pool were sequenced. For ToBRFV infection in Tm‐2 ² plants, progeny sequencing was performed at 10 and 20 dpi, with only the first three infected plants included. For Tm‐2 ² plant 1, 12 and 11 clones were sequenced at 10 and 20 dpi, respectively. At 10 dpi, 8 clones encoded tyrosine (Y) and 4 clones encoded histidine (H); at 20 dpi, all 11 clones encoded tyrosine (Y). For Tm‐2 ² plant 2, 12 and 12 clones were sequenced at 10 and 20 dpi, respectively, and all clones encoded tyrosine (Y). For Tm‐2 ² plant 3, 10 and 10 clones were sequenced at 10 and 20 dpi, respectively, and all clones encoded tyrosine (Y). For ToBRFV infection in WT plants, sequencing was performed only at 10 dpi, as the plants had died by 20 dpi. A total of 12, 10 and 12 clones were sequenced for plants 1–3, and all clones encoded tryptophan (W) (Table 3). Progeny of ToBRFV H67C in WT plants were also sequenced and the introduced mutation was maintained.

TABLE 3.

Progeny sequences of tomato brown rugose fruit virus (ToBRFV) isolated from Tm‐2 ²‐transgenic and wild‐type (WT) Nicotiana benthamiana infected with ToBRFV movement protein (MP) H67 mutant pool.

Time (days post‐inoculation)	Tm‐2 2			WT
	Plant 1	Plant 2	Plant 3	Plant 1	Plant 2	Plant 3
10	8 TAT (Y), 4 CAC (H)	12 TAC (Y)	10 TAC (Y)	12 TGG (W)	10 TGG (W)	12 TGG (W)
20	11 TAT (Y)	12 TAC (Y)	10 TAC (Y)	NA	NA	NA

Note: The number of clones and the genome sequences at the H67 position are presented, with the corresponding amino acid sequence highlighted in red.

Abbreviation: NA, not available.

TABLE 4.

Progeny sequences of tobacco mosaic virus (TMV) isolated from wild‐type (WT) Nicotiana benthamiana infected with TMV movement protein (MP) C68 mutant pool.

Time (days post‐inoculation)	WT
	Plant 1	Plant 2	Plant 3
10	10 ACT (T)	10 ACT (T)	8 ACT (T), 2 ACA (T)

Note: The number of clones and the genome sequences at the C68 position are presented, with the corresponding amino acid sequence highlighted in red.

Progeny sequences of TMV isolated from two Tm‐2 ² and WT N. benthamiana plants infected with the TMV MP C68 mutant pool were sequenced at 10 dpi, as all plants were dead by 20 dpi. A total of 10 clones were sequenced for each plant, and all clones encoded threonine (T). Two codons, ACT and ACA, were observed in plant 3, whereas only ACT was detected in the other plants (Table 4). Progeny of TMV C68H in WT plants were also sequenced and the introduced mutation was maintained.

2.3. Single‐Infection and Co‐Infection Assays Identify Infectious Single Mutants With Lower Competitive Fitness Than the WT

In the mutant pool infection assays, the ToBRFV MP H67 TAC (Y) mutant was detected in JP tomato plants and Tm‐2 ² N. benthamiana plants, whereas CAC (H), TTT (F) and TTC (F) mutants were frequently observed in RG plants; TGG (W) was frequently observed in RG tomato and WT N. benthamiana plants (Tables 1 and 3). Notably, although CAC encodes histidine (H), this codon differs from the WT CAT and was additionally detected in Tm‐2 ² N. benthamiana plant 1 (Table 3). In the Solanaceae‐infecting subgroup, the position corresponding to H67 in ToBRFV MP is occupied by asparagine (N) in YTMMV and is encoded by AAT (Figure 1a). Based on these results, these six mutants were selected for further infectivity assays in JP and RG plants. In addition, co‐infection with the WT virus was performed to assess their potential to outcompete the WT.

Four tomato plants were included in each experiment, and infection was assessed by western blot analysis of ToBRFV CP at 4 wpi. At 4 wpi, in JP plants, the infection rates of the TAC (Y), CAC (H), TTT (F), TTC (F), AAT (N) and TGG (W) mutants were 4/4, 0/4, 2/4, 1/4, 4/4 and 4/4, respectively, although their accumulation levels were markedly lower than that of the WT virus (Figure 3a). In RG plants, the corresponding infection rates were 4/4, 4/4, 2/4, 3/4, 4/4 and 4/4, respectively (Figure 3a). In the co‐infection assays, infection was detected in all plants (4/4) for all combinations, except for WT + TTC in RG plants, where the infection rate was 3/4 (Figure 3a). Symptoms were recorded at 6 wpi (Figure 3b). This experiment was repeated once. In this repeated experiment, infection was detected at both 2 and 4 wpi. At 2 wpi, in JP plants, only TTC (F) and AAT (N) mutants were detected, and infection rates were 4/4 and 1/4, respectively. In RG plants, the infection rates of TTT (F), TTC (F) and AAT (N) mutants were 4/4, 4/4 and 3/4, respectively. Other mutants were not detected (Figure S3a). In the co‐infection assays, infection was detected in all plants (4/4) for all combinations, except for WT + CAC (H) in JP plants, where the infection rate was 2/4 (Figure S3a). At 4 wpi, the infection rates of the TAC (Y), CAC (H), TGG (W) and AAT (N) mutants were mostly the same as the first experiment in both JP and RG plants, whereas infection rates of TTT (F) and TTC (F) increased to 4/4 in JP plants (Figure S3b). In all co‐infection groups, an infection rate of 4/4 was reached in all cases. Detailed symptoms of plants used in the second experiment were shown in Figure S4.

FIGURE 3.

Analysis of tomato brown rugose fruit virus (ToBRFV) infection in Jingpeng No. 1 (JP) and Rutgers (RG) tomato plants either infected with single mutants or co‐infected with wild type (WT) and single mutants. (a) Western blot analysis of the infection at 4 weeks post‐inoculation (wpi). (b) Symptoms of JP and RG tomato plants infected with ToBRFV at 6 wpi. Scale bar = 7 cm. CBB, Coomassie Brilliant Blue loading control; CP, coat protein; M, mock; P, positive control; RBL, RuBisCO loading control.

Symptoms and necrosis were scored according to previous studies (Bock et al. 2022; González‐Concha et al. 2023). Disease severity index (DSI%) and necrosis index (necrosis%) were calculated. For tomato plants, DSI% was generally lower in JP than in RG (Table S3). In contrast, necrosis% was higher in JP plants than in RG plants; in fact, necrosis% was 0 in all RG plants (Table S4).

Progeny sequencing was performed for the first three plants in both experiments when the infection rate exceeded 3/4. When the infection rate was below 3/4, all infected plants were included. In each case, up to six plants were included. Two or three clones were sequenced per plant. At 2 wpi, in both JP and RG plants, TTT (F) and AAT (N) reverted to the WT sequence, whereas for the TTC (F) mutant, the introduced mutation was maintained in both plants (Table S5). In all co‐infection assays, only the WT sequence was detected, except in the WT + TTC (F) treatment, where the introduced mutation was maintained in one JP plant (Table S5). At 4 wpi, for the TAC (Y) mutant, the introduced mutation was maintained in all RG plants but reverted to the WT sequence in all JP plants. In both JP and RG plants, the CAC (H) and TTT (F) mutants consistently reverted to WT. Notably, the introduced mutation in the TTC (F) mutant was maintained in all JP and RG plants. For the TGG (W) mutant infection in JP plants, three clones from plant 1 and three clones from plant 5 reverted to the WT sequence. In plant 2, three clones maintained the introduced mutation. In plant 3, TTG (L) and GGT (G) were detected in two and one clones, respectively. In plant 4, two clones maintained the introduced mutation, and TTT (F) was detected in the third clone. In plant 6, one clone reverted to the WT sequence, one clone maintained the introduced mutation, and CCT (P) was detected in another clone (Table S6). In RG plant 1, 2, 4 and 5, all clones were TGC (C). In plants 3 and 6, all clones maintained the introduced mutation. In all co‐infection assays, only the WT sequence was detected, except that in plants 4–6 of the WT (H) + TTC (F) group, where only TTC (F) was detected (Table S6). These results suggest that although these mutants were detected in the mutant pool infections, most were likely carried by other mutants, as they failed to stably infect JP and/or RG plants. Although TTC (F) was able to infect both JP and RG plants, it was outcompeted by WT during co‐infection in three out of six cases. In one case, TTC (F) outcompeted WT. These data suggest that TTC likely has lower competitive fitness relative to WT, but stochastic selection may also play a role.

The single‐infection and co‐infection assays were also repeated in WT and Tm‐2 ² N. benthamiana plants, with three plants included in each treatment. Infection was assessed at 10 dpi by western blot analysis of CP. The infection rate was 3/3 in all cases, except that the TTC (F) and AAT (N) mutants failed to infect any Tm‐2 ² N. benthamiana plants (Figure 4a). Symptoms were recorded at 20 dpi (Figure 4b). Only symptom scoring was performed on N. benthamiana plants because N. benthamiana plants collapsed in most cases. Symptom scoring showed that DSI% was generally lower in Tm‐2 ² than in WT plants (Table S7). This N. benthamiana experiment was also repeated once; the same results were obtained (Figure S5).

FIGURE 4.

Analysis of tomato brown rugose fruit virus (ToBRFV) infection in wild‐type (WT) Nicotiana benthamiana and N. benthamiana with Tm‐2 ² transgene either infected with single mutants or co‐infected with WT and single mutants. (a) Western blot analysis of the infection at 10 days post‐inoculation (dpi). (b) Symptoms in WT N. benthamiana and N. benthamiana with the Tm‐2 ² transgene, infected with ToBRFV at 20 dpi. CBB, Coomassie Brilliant Blue loading control; CP, coat protein; M, mock; P, positive control; RBL, RuBisCO loading control.

Progeny sequencing was performed for all infected plants from these two experiments. For the TAC (Y) mutant, the introduced mutation was stably maintained in both WT and Tm‐2 ² plants (Table S8). This contrasts with the results observed in tomato plants, where TAC (Y) failed to stably infect JP plants (Tm‐2 ² /tm‐2). These findings suggest that genetic factors other than Tm‐2 ² in JP plants may contribute to the reversion of TAC (Y) to the WT sequence. For the CAC (H) mutant, reversion to the WT sequence was observed in most cases; however, in some WT and Tm‐2 ² plants, TAC (Y) was detected. For the TTT (F) and AAT (N) mutant, all progeny reverted to WT. The behaviour of the TTC (F) mutant was more complex—in Tm‐2 ² plants, most cases reverted to WT, with TAC (Y) detected in one case, and TTC (F) mutation was maintained in two plants. In contrast, in WT plants, the introduced TTC (F) mutation was maintained in all cases. For the TGG mutant, the introduced mutation was stably maintained in both WT and Tm‐2 ² plants. In all co‐infection assays except Tm‐2 ² plants of the WT + TGG group, only the WT sequence was detected (Table S8). In Tm‐2 ² plants of the WT + TGG group, the TGG (W) mutation was maintained in all clones from plants 1 and 4. In plant 2, TGG (W), TGC (C) and TTC (F) were each detected once. In plant 3, two TGG (W) and one WT (H) were detected. In plant 5, two TGG (W) and one TGC (C) were detected. In plant 6, two WT (H) and one TGG (W) were detected (Table S8).

Additionally, the complete genome of the TTC (F) mutant recovered from an infected N. benthamiana Tm‐2 ² transgenic plant (plant No. 6) was amplified to determine whether infectivity was associated with additional genomic mutations. One clone was included and the sequence has been deposited in GenBank under accession number PZ420571. When the progeny sequence was compared with the TTC (F) sequence, 10 nucleotide substitutions were identified: eight in the 183 kDa replicase coding region, one in the MP coding region, and one in the CP coding region. Among the eight substitutions within the 183 kDa replicase coding region, seven were non‐synonymous, resulting in the following amino acid changes: isoleucine to threonine at position 15 (I15T), threonine to serine at position 489 (T489S), lysine to arginine at position 593 (K593R), arginine to lysine at position 793 (R793K), arginine to glutamine at position 1079 (R1079Q), serine to cysteine at position 1262 (S1262C) and aspartic acid to glutamic acid at position 1291 (D1291E). The remaining substitution, at position 2971, was synonymous. In the MP coding region, a substitution at position 5145 changed codon 80 from CCA to TCA, resulting in a proline‐to‐serine change (P80S). In the CP coding region, a substitution at position 5971 changed codon 87 from GCT to GCC. This change was synonymous; therefore, no amino acid alteration was detected in the CP (File S1 and Table S9).

2.4. Local RNA Secondary Structure and Altered Stability of the Tm‐2 ² –ToBRFV Complex Caused by the Aromaticity at Position 67 Are Unlikely to Contribute to the Escape of ToBRFV From Tm‐2 ² ‐Mediated Resistance

The secondary structures of the genomic RNA of the WT and six mutants used in Figures 3 and 4 (single infection and co‐infection with WT) were predicted using RNAfold. Both infectious and non‐infectious mutants failed to maintain the WT‐like local secondary structure (Figure S6). Therefore, the local RNA structure is unlikely to be responsible for the escape of ToBRFV from Tm‐2 ² ‐mediated resistance.

To clarify the role of the aromatic side chain at residue 67 of the ToBRFV MP in infection, PDBePISA (https://www.ebi.ac.uk/pdbe/pisa/) was used to analyse the interfacial characteristics of mutants at this position in the predicted complex with Tm‐2 ² . Compared with non‐infectious mutants, infectious mutants did not exhibit consistently higher or lower ΔⁱG values or interface areas (Table S10). These data suggest that the aromatic nature at this position does not facilitate escape from Tm‐2 ² recognition by reducing complex stability. We speculate that a third protein or small ligand may be involved in mediating the interaction between Tm‐2 ² and MP through this conserved position.

3. Discussion

This study highlights the essential role of the aromatic nature of residue 67 in the ToBRFV MP in evading Tm‐2 ² ‐mediated resistance. These findings lay the groundwork for future research focused on developing novel strategies to confer resistance to ToBRFV.

A key finding of this study is that the RNA motif at the H67 position of ToBRFV plays a crucial role in its infection process. Although CAC encodes the WT histidine (H), it differs from the WT (CAT). Infection assays in both tomato and N. benthamiana (Tables S5, S6 and S8) revealed that CAC (H) failed to support infection. Moreover, TTC (F), but not TTT (F), infected JP tomato. In silico prediction of the RNA secondary structures of the genomic RNA was performed for all mutants used in both single‐infection and co‐infection assays. The results showed that both the infectious mutants and those that failed to maintain the secondary structure exhibited altered local structures (Figure S6). This suggests that other features of this nucleotide sequence, such as codon usage, may be pivotal for ToBRFV infection. These observations are consistent with recent codon‐level analyses that highlight codon usage bias and broader evolutionary constraints in ToBRFV (Ghorbani 2024). This is in line with studies on viroids, which, lacking protein‐coding ability, rely entirely on sequence, RNA structures, and interactions with host factors for infection (Wu and Bisaro 2024a, 2025; Wu et al. 2019, 2024, 2020). Future studies should explore the function of the MP RNA.

Mutant pool infection assays revealed that the aromatic mutants TAC (Y) and TAT (Y), along with CAC, which encodes the WT histidine (H), were detected in JP tomato and Tm‐2 ² N. benthamiana. In single infection assay, ToBRFV carrying the aromatic TTC (F) and TGG (W) at the 67th position infected JP tomato and Tm‐2 ² N. benthamiana, while the aromatic TAC (Y) enabled infection of Tm‐2 ² N. benthamiana. Across multiple biological replicates and in two host species, we consistently observed enrichment of aromatic residues at this position under Tm‐2 ² selection. Functional analyses further showed that non‐aromatic substitutions were unable to infect Tm‐2 ² plants, whereas aromatic substitutions retained infectivity. These results suggest that the aromatic moiety at this position, rather than specific side‐chain polarity or charge, is crucial for evading Tm‐2 ² ‐mediated resistance, aligning with the imidazole ring of the WT histidine, which also exhibits aromatic characteristics. The five‐membered heterocyclic ring structure found in the side chain of the amino acid histidine possesses a specific set of chemical properties consistent with aromatic compounds. These properties include a planar structure, a cyclic arrangement of atoms, and a delocalized system of π electrons, which collectively contribute to enhanced stability (Calinsky and Levy 2024; Sundberg and Martin 1974). PDBePISA analysis of residue 67 mutants in the predicted Tm‐2 ² –MP complex revealed no consistent differences in ΔⁱG or interface area between infectious and non‐infectious mutants (Table S10). These results suggest that aromaticity at this position does not promote escape from Tm‐2 ² recognition by altering complex stability. Instead, it is possible that a third protein or small ligand mediates the interaction between Tm‐2 ² and MP via this conserved site.

Interestingly, although TGG (W) was not accumulated to detectable levels in the mutant pool infection assay in both species carrying the Tm‐2 ² gene, it was detected in RG tomato at 2 wpi. In addition, this TGG (W) was also detected in WT N. benthamiana (Tables 1 and 3). In the single infection assay, the TGG (W) mutant was capable of infecting tomato and N. benthamiana plants carrying the Tm‐2 ² resistance gene. In the co‐infection assay with WT, the TGG (W) mutant even outcompeted WT in Tm‐2 ² transgenic N. benthamiana plants (Tables S6 and S8). We speculate that the interaction between mutants in the pool infection may create an unfavourable condition for the accumulation of TGG (W). This finding reflects the complexity of the initial viral sequence population, species and resistance gene‐determined viral accumulation.

Comparing the pooled infection and single mutant infection assays, we found that the five single mutants, TAC (Y), CAC (H), TTT (F), TTC (F) and TGG (W) persisted in pooled infections (Tables 1 and 3), yet failed in most single infection assays (Tables S5, S6 and S8). These results suggest that the presence of functional mutants or WT in the same population may facilitate the maintenance of these mutants. They are consistent with the context‐dependent detectability of less‐fit genotypes in co‐infection versus single infections (Taylor et al. 2025; Wu and Bisaro 2024b).

Interestingly, a recent study demonstrated that replacing the nucleotide sequence encoding TMV MP C68 with the ToBRFV MP H67 codon (CAT) rendered the virus non‐infectious (Hak et al. 2023). In contrast, our study found that TMV MP C68 failed to infect JP tomato and Tm‐2 ² N. benthamiana, but successfully infected RG tomato and WT N. benthamiana. These results suggest that simply replacing TMV MP C68 with histidine is insufficient to evade Tm‐2 ² ‐mediated resistance, indicating that additional factors may be involved. Future research should explore the discrepancy between our findings and the previous study.

In conclusion, this study provides evidence that the aromatic nature of residue 67 in the ToBRFV MP is probably a critical determinant in evading Tm‐2 ² ‐mediated resistance. The selective pressure for aromatic residues in Tm‐2 ² ‐resistant backgrounds, combined with the observed fitness costs associated with these evasion mutations, enhances our understanding of ToBRFV pathogenesis. These findings offer valuable insights for the development of resistance strategies against ToBRFV by engineering the Tm‐2 ² gene to recognize the conserved aromatic side chain. Further investigation into the molecular interactions mediated by this aromatic residue, as well as other genetic factors influencing resistance in tomato cultivars, is warranted to refine these strategies.

4. Experimental Procedures

4.1. Preparation of Mutant Pools and Vector Construction

TMV MP C68 and ToBRFV MP H67 mutant pools were generated by replacing the target sequences in the pCB301 vector, which expresses TMV and ToBRFV cDNA. Infectious clones of the TMV and ToBRFV strains were generated in our laboratory (Zhang et al. 2023). Briefly, to generate the mutant pools, two PCRs were performed. One PCR used a pair of normal primers, while the other PCR used a degenerate forward primer covering the target region and a normal reverse primer. The resulting PCR products were subjected to homologous recombination to construct the mutant pools. The TMV MP C68 mutant pool construct was prepared using TMV as the template, with homologous recombination carried out on two sequences: TMV‐C68poolBackone1 (primers: T‐F/TMV‐51R) and TMV‐C68poolBackone2 (primers: TMV‐C68NNN‐F/T‐R). The ToBRFV MP H67 mutant pool construct was prepared by homologous recombination of two sequences: ToBRFV‐Backone1 (primers: T‐F/ToBRFV‐51R) and ToBRFV‐Backone2 (primers: ToBRFV‐H67NNN‐F/T‐R). The homologous recombination products were used to transform Escherichia coli XL10 competent cells. More than 10‐fold clones (> 640) were collected and pooled. These libraries were subjected to deep sequencing on an Illumina NovaSeq 6000 platform. More detailed methods are described in a previous study (Wu et al. 2024).

Two single mutants, ToBRFV MP H67C and TMV MP C68H, were constructed by replacing the nucleotide sequence encoding the 67th residue (H67) of the ToBRFV MP with the nucleotide sequence encoding the 68th residue (C68) of the TMV MP, and vice versa for the TMV MP C68H mutant. The primer pairs used in the first PCR were T‐F/TMV‐51R and T‐F/ToBRFV‐51R, and those used in the second PCR were TMV‐C68H‐F/T‐R and ToBRFV‐H67C‐F/T‐R. The resulting PCR products were subjected to homologous recombination to generate the mutants.

The pCB301 vector expressing ToBRFV cDNA was also used to prepare four single mutants: ToBRFV MP TAC (Y), CAC (H), TTT (F) and TTC (F). ToBRFV MP TAC (Y) was constructed by introducing two mutations, C5106U and T5108C. Two PCRs were conducted: the primer pair T‐F/ToBRFV‐51R was used in the first PCR, and the primer pair ToBRFV‐TAC‐F/T‐R was used in the second PCR. Homologous recombination was then performed. ToBRFV MP CAC (H) was constructed through site‐directed mutagenesis using the primer pair ToBRFV‐CAC‐F/R, introducing a single mutation, T5106C. ToBRFV MP TTT (F) was prepared by introducing two mutations, C5106U and A5107U. The primer pair T‐F/ToBRFV‐51R was used in the first PCR, and the primer pair ToBRFV‐TTT‐F/T‐R was used in the second PCR. Homologous recombination was then performed. ToBRFV MP TTC (F) was prepared by introducing three mutations: C5106U, A5107U and U5108C. Two PCRs were conducted: the primer pair T‐F/ToBRFV‐51R was used in the first PCR, and the primer pair ToBRFV‐TTC‐F/T‐R was used in the second PCR. Homologous recombination was then performed. Primers are listed in Table S11.

4.2. Plants and Agroinfiltration

Tomato cultivars JP (Tm‐2 ² /tm‐2) or RG (tm‐2/tm‐2) and N. benthamiana plants were grown at 25°C with a 16‐h light/8‐h dark cycle. JP seeds were sourced from Xi'an Jinpeng Seedling Company, and Tm‐2 ² ‐transgenic N. benthamiana was kindly provided by Professor Yule Liu at Tsinghua University. Rutgers and WT N. benthamiana plants were grown in‐house. Agrobacterium tumefaciens C58C1‐mediated infection was performed at the two‐cotyledon stage for tomato plants and at the first two fully expanded true leaves of 3‐week‐old N. benthamiana plants. Agrobacterium cultures containing viral clones were infiltrated into leaves at an OD₆₀₀ of 0.5. For co‐infection, Agrobacterium cultures of each strain were adjusted to an OD₆₀₀ of 1.0 and mixed in a 1:1 ratio to achieve a final OD₆₀₀ of 0.5 for each. Systemic leaves of tomato were collected at 2, 4 and 6 wpi for reverse transcription (RT)‐PCR and western blot analysis. Systemic leaves of N. benthamiana samples were collected at 10 dpi for RT‐PCR and western blot analysis.

4.3. Protein Preparation and Western Blot

To assess TMV and ToBRFV CP accumulation in systemic leaves, total protein extracts were prepared as previously described (Han et al. 2023). Proteins (extracted from ~500 ng of leaf tissue) were separated by SDS‐PAGE on 12% gels and transferred to methanol‐activated PVDF or 0.2 μm nitrocellulose membranes. Mouse anti‐ToBRFV CP antibody (1:8000) and mouse anti‐TMV CP antibody (1:8000) were used as primary antibodies, followed by goat anti‐mouse IgG secondary antibody (1:10,000; Transgen). Signals were visualized using the Invitrogen iBright 1500 imaging system.

4.4. RNA Preparation, RT‐PCR and RNA Sequencing Analysis

Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific) following the manufacturer's protocol. RT‐PCR was performed to verify the presence of viral genome RNA in the systemic leaves of each experimental group using the primers ToBRFV‐dec‐F/R or TMV‐dec‐F/R. Reverse transcription (RT) was carried out using ReverTra Ace (Toyobo) with random primers and oligo(dT) to synthesize the first cDNA. The cDNA was used as templates for PCR amplification with KOD‐FX‐Neo (Toyobo). Sanger sequencing was performed to assess the maintenance of introduced mutations. All primers are listed in Table S11.

4.5. Symptom Scoring and Infection Assessment

Disease severity index (DSI) was used to assess the severity of symptoms caused by ToBRFV and mutants in tomato and N. benthamiana. The method was described in previous studies (Bock et al. 2022; González‐Concha et al. 2023). Symptoms were scored on a 0–5 scale: 0 = no visible symptoms; 1 = slight curling; 2 = mild curling, stunting or mosaic; 3 = curling, stunting, mosaic or chlorosis; 4 = curling with marked leaf narrowing and size reduction, accompanied by strong stunting, chlorosis or necrosis; and 5 = plant decline or near collapse with very severe deformation and stunting. The sum and median values of severity were calculated based on the symptom scores. DSI% was calculated using the following formula: DSI (%) = Sum/(N × 5) × 100, where Sum is the sum severity and N is the number of plants. Lower‐ and middle‐leaf necrosis was scored on a 0–5 scale: 0 = no necrosis (0% leaf area); 1 = a few necrotic spots (≤ 5% leaf area); 2 = mild necrosis (> 5%–15% leaf area); 3 = moderate, coalescing necrotic patches (> 15%–30% leaf area); 4 = severe, extensive necrosis and tissue collapse (> 30%–70% leaf area); and 5 = widespread or systemic necrosis with strong plant decline (> 70% leaf area or near‐lethal decline). The necrosis scoring was performed only on the middle and lower leaves of tomato plants, because N. benthamiana plants collapsed in most cases. Sum and median necrosis scores were calculated. Necrosis% was calculated using the following formula: Necrosis (%) = (Sum/(N × 5) × 100), where Sum is the sum necrosis and N is the number of plants.

Author Contributions

Luyou Wang: investigation. Yuhong Zhang: methodology, investigation, formal analysis, writing – original draft, visualization. Jianping Chen: formal analysis, writing – review and editing, resources. Jiayi Liu: investigation. Yuxin Nie: investigation. Fei Yan: formal analysis, writing – review and editing, funding, resources. Junchen Xu: investigation. Jian Wu: conceptualization, investigation, formal analysis, writing – review and editing, project administration, funding acquisition, methodology, supervision, resources, data curation, visualization, validation.

Funding

This work was supported by the National Key R & D Program of China (2022YFD1401200) and the National Natural Science Foundation of China (32470149, 32272483).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Detailed disease symptoms in tomato plants infected with the ToBRFV H67 pool, TMV C68 pool, ToBRFV H67C (CAT → TGT) and TMV C68H (TGT → CAT). The nucleotide changes in each mutant are indicated in parentheses. This figure provides supplementary information for Figure 1e.

Figure S2: A repeated experiments for Figure 1c ToBRFV H67 pool and ToBRFV H67C infection in Tm‐2 ² tomato plants. See Figure 1c for details. Here infection was detected at 2 weeks post‐inoculation.

Figure S3: A repeated experiments for Figure 3. See Figure 3 for details. Here infection was detected at both (a) 2 weeks post‐inoculation (wpi) and (b) 4 wpi. M, mock; P, positive control.

Figure S4: Detailed disease symptoms in ToBRFV single mutants infected Jingpeng No. 1 (JP) and Rutgers (RG) plants. This is supplementary information for Figure 3b.

Figure S5: A repeated experiments for Figure 4. See Figure 4 for details. Here infection was detected at 10 days post‐inoculation.

Figure S6: Changes in RNA secondary structure of genomic RNA caused by mutations used in Figures 3 and 4. The secondary structures of the genomic RNA of the wild type (WT) and six mutants used in Figures 3 and 4 (single infection and co‐infection with WT) were predicted using RNAfold. Local structure was presented.

File S1: Complete genomic nucleotide sequence alignment of the TTC (F) mutant with the wild‐type (WT) sequence. Red frames indicate the mutated sites.

Table S1: Sources and characteristics of nucleotide sequences from 32 tobamoviruses.

Table S2: Read counts for each codon in the ToBRFV MP H67 pool and TMV MP C68 pool.

Table S3: Symptom scoring and infection assessment of tomato systemic leaves.

Table S4: Necrosis scoring of tomato lower‐middle leaves.

Table S5: Progeny sequences of ToBRFV isolated from Jingpeng No. 1 (JP) and Rutgers (RG) tomato cultivars at 2 weeks post‐inoculation, either infected with ToBRFV MP single mutants or coinfected with wild‐type and single mutants.

Table S6: Progeny sequences of ToBRFV isolated from Jingpeng No. 1 (JP) and Rutgers (RG) tomato cultivars at 4 weeks post‐inoculation, either infected with ToBRFV MP single mutants or coinfected with wild‐type and single mutants.

Table S7: Symptom scoring and infection assessment in Nicotiana benthamiana plants.

Table S8: Progeny sequences of ToBRFV isolated from Tm‐2 ² and WT two Nicotiana benthamiana genotypes, either infected with ToBRFV MP single mutants or coinfected with wild‐type and single mutants.

Table S9: Nucleotide substitutions and corresponding amino acid changes identified in the complete genome of the ToBRFV TTC (F) mutant recovered from an infected Nicotiana benthamiana Tm‐2 ² transgenic plant.

Table S10: PDBePISA‐derived interface parameters of predicted complexes between Tm‐2 ² and MP residue 67 variants.

Table S11: Primers used in this study.

Data Availability Statement

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